Amyotrophie Lateral Selves
Quantitative Assessment of
Neuromuscular Deficit in ALS
Patricia L. Andres, M.S, P.T..* Linda M. Thibodeau, P-.A.,¢
Loren: J. Finison, Ph-D.,4 and Theodove 1. Munsat, M.D.§
Accurate measurement is as essential in clinical research as it is in
basic research. The valiity of data resulting from an investigation depends
in great part on the measurement tools avaihble, Each year hundreds of
thousands of dollars are spent on dinial trials in neuromuseu
using measurement techniques of questionable accuraey.. Curt
‘of measuring neuromuscular deficit include subjective elements and Tack
the sensitivity to detect small but meaningful change in deterioration ancl
therapeutic efficacy. Accirate, objective measurement tools are needed to
document the natural hstory of progressive diseases, such as ALS, and
thus provide clues to etiology and pathogenesis.”
ALS is a disease of progressive damage to upper motor neurons and
lower motor neurons resilting in the loss of voluntary motor function and
subsequent the loss of functional capacity.*** Despite major technologie
advances in our ability todiagnose specific neuromuscular diseases, current
methods to quantitate de‘cit in ALS remain primitive, Unfortunately, there
are no precise chemical markers or electrophysiologic techniques available
to document disease progression. Therefore, motor neuron loss must be
measured by indirect methods,
‘This article describes various techniques used to measure strength and
function in ALS, particularly our testing protocol, the Tufts Quantitative
Lecturer ta Newslony
‘Reveuch Physi Therapie stant
{Asuodate Clieal Profesor, Department of Comunity Health; Statistica Consultant
iDimetor, Nouromeiar Research Unt Prfeor of Nerology
Frum te Neuromuscular Research, Unit Department of Neurology, Tult-New England
‘Medical Center, Baton, Massachnstis,
Supported by grams fiom the Muscular Dystophy Assodaton and NIH Grant CRC No,
‘MoaRRiebes
Nenroloie Cline Va. 8, Na. 1, Fey 1957 135136 Parmicia Le ANDRES ET AL
Neuromuscular Exam. It will aso address basie considerations in designing
test batteries for measuring deficit in neuromuscular diseases,
SUMMARY OF CURRENT METHODS
‘Traditional methods of evaluating ALS patients include the clinical
neurologic exam, manual musele test (MMT),*" 2" rating seales,**"® and
timed functional activites.* ** MMT and ‘rating scales: produce ordinal
data. That is, categories are arranged in ranked order, but intervals between
rankings are not eaval. For example, spasticity can be described as L: mild
2: moderate, or 3s severe. Muscles in the mild category have less spasticity
than those in the moderate category, but the difference between mild andl
‘moderate is not necessarily the same as between moderate and severe, In
addition, a muscle with mild spasticity is not equal to all other muscles
graded mild, Standard statistical analyses based on means are not legit
imately applied to ordinal data, as the individual scores are not evenly
spaced. Ordinal data permits only analysis of Frequencies.*
Manual Muscle Test
MMT, the most commonly used test of muscle strength, was originally
developed’ by Lovett to evaluate isolated weakness associated with polion
iyelits It is a standardized grading system in which strength is subjec-
lively graded according to the patients ability to move against gravity and
hold agsinst the examiner's manual resistance." Several groups, including
the Medical Research Counecil,® Kendall and Kendall," and Daniels and
Worthingham,” have devised variations on the MMT technique
MMT is deficient in several respects. Its quantitative capacity is limited
because ordinal data is produced. Grading is subjective, and scoring is
influenced by the examiners strength and the weight of the patient's Timb,
Finally, the number of grades is Hinited,
Beasley* demonstrated the deficiencies of MMT by comparing it sith
‘quantitative isometric force measurement of both normal subjects
‘with neuromuscular delieit, He showed thatthe force generated by muscles
graded in the same MMT grade can vary as much as 40 per cent. As a
‘muscle remains in the same grade until it reaches a “eritical” level, the use
Of MMT in a longitudinal stady may falsely indicate plateaning or stepwise
progression when, in fact, strength is declining in a Hinear manner
Functional Rating Seales
Functional seales are widely used in assessing neuromuscular deficit
21.40 They are easily performed and broadly reflect clinically relevant
functional disturbances. However, they are relatively insensitive to the
degree of motor neuron loss; they produce ordinal data and rely on the
subjective judgment of the examiner. Functional scales can provide such
formation as the number of patients who improve or worsen, but they do
not show how much or how fist deterioration takes place
'e most commonly used funtion seale in assessing ALS deficit is
(LS Score” designed by Norrs.™ It consists of 34 items rated 0 to 3
the "Quavrtrarive Assessuenr oF Neunowuscutan Drricrr ALS 137
(0 = no use, L = very limited, 2 = useful but impaired, 3
‘normal composite seore i 100. The items include functional activities and
neurologic signs such as chewing, rolling, feeding, breathing, reflexes, and
plantar responses, As subjective reflex testing weighted equally with
Such essential finetions as breathing and swallowing, the score is a doubtful
index of disease progression. In addition, the sensitivity of the seale is
limited because only fonr grades are used and thus fewer options are
available. For example, compromised breathing must be either graded
tusefal but impaired (2) or very limited (1). Finally, no reliability or validity
studies of the ALS Score have been published.
‘Timed Functional Activities
Several investigators have included timed functional activites in their
test batteries to measure neuromuscular deficit * * This approach
Produces interval data that can be analyzed using parametric statistics.
‘Avo, timed tests are more objective and reduce examiner bis.
“Timed tests reflect impaired physical function and ean providea general
indication of improvement of deterioration. However, they may not always
provide the best means of charting disease progression. Certain tasks may
Fequire a eritieal number of motor units and the time required for theit
performance may not cage until this evita level is reached. Figure 1
lustrates the decline in arm strength (maximal isometric force) eompared
to the decline in the ably 10 dial a phone number and place pexs in a
Soar sis Ut long sepa detente in nar bho,
timed hand activities Ampped of sharply at a certain point. In addition,
timed activites may be infuenced by environmental or other factors
texteral 0 the disease proces. Figure 2 shows that timed ambulation, for
instance, may actually improve dramatically when tho typo of assistance is
changed.
4.57 tived nand function and arn strength over tine
“
seo 9601000 060 1100 1160 120012501300,
ays since disease coset
Figure 1. In this patient ere & « contauous ines decline in arm atensth while timed
han neton dopa of sharply er approsinatly dy 110,138, Partita Le ANDRES EA
no alo abs she alo sho ohn obo so
sags sine sisease Oat
igure 2 thi raph teats ted wal et of single patient over te. The pase
‘anbslated independent unt day 100 err, when he started sing x alg walker, Then
ter day 00 (arrow) the patent eld no Tonger se the walker and depended pasa
tstctance to ambulte,
Several investigators have reported the use of timed Funetional activities
to measure change in ALS drug trial." * Timed activities may be of
eater value in acute studies than in longitudinal studies because environ-
‘mental fitrs and eral levels are uniely to influence the fet of the
ng
‘Quantitative Strength Measurement
Strength is dependent upon many factors including type of contraction
isometric, isotonic, isokinetic), speed of contraction, length-tension rela-
tionship, neuronal discharge, and cross-sectional area of muscle and moti-
vation.” Recently developed quantitative methods to measure strength
Involve either isokinetic oF isometric testing
‘An isokinetic dynamometer (for example, Cybex 11* and Kin-Comf)
contains a preset speed-controlling mechanism that offers resistance pro-
portional to the dynamic tension developed by the exereising limb." The
torque produced by a limb as it moves throughout the range of motion is,
recorded. The device produces accurate interval readouts andl measures
torque throughout the entire range of motion while controlling velocity
High test-retest reliability was demonstrated in children testing 10 large
muscle groups. Isokinetic dynamometers have several. disadvantages,
however. Their cost may be prohibitive at $25,000 to $35,000, Cybex
adjustment is difficult, making measurement of several diferent muscle
groups laborious. Also, muscles with less than Fair-+ strength cannot be
tested through the full are of motion because the limb is too weak to exert
force against the machine's moving lever arm,
‘There are several instruments for quantitating maximal voluntary
Isometric force, ‘These include spring balances, cable tensiometers, hy
“Gybex Dison of Lumex, In, 2100 Smithtown Ave, Rovonkoms, NY
Chater Corp, 101 Mem Dr, BO Box 4287, Catan, IN‘Quavrexrive Assessunr ov NEUROMUSCULAR Devicrr i ALS 129
raulie myometers, and various types of strain gauges." Isometric testing
has the advantage of eliminating velocity and muscle length as variables.”
2304 However, the examiner must be carefully trained to assure that the
limb position and stabilization are standard.""®
Hand-held myometers and electronic strain gauges are widely used to
{quantitate isometric force. Hand-held myometers have been used in studies
{0 measure neuromuscular deficit, yiekling exellent test-retest reliability.”
‘Their pocket size and portability ae advantageous, but a major disadvantage
is that the maximum force recorded is limited by the strength of the
texaminer, who must be able to hold against the patients resistance. The
hand-held myometer is therefore not suitable for measuring minor degrees
of weakness in large adulls. Also, it does not produce a hard copy of results
Several Investigators have used strain-gauges to measure isometric
force." ® Their main advantages are that (1) a hard copy of the test
results is produced and (2) the subject is pulling against an immovable
rap, rather than against the force of the examiner (Fig. 2)
DEVELOPMENT OF THE TUFTS QUANTITATIVE
NEUROMUSCULAR EXAM
We have developed a standardized test battery to quantitate neuro:
muscular deficit in ALS, the Tufts Quantitative Neuromuscular Exam
Figute 3. metric ree measurements performed on a standard ection table
with aminaenoprights and aijostble ngs A'spon strap poxtoned on the lib ne
Table Yor devas on pining The stp i connected tothe san gige, whi
gon a nme upright The strap and ral gauge ema pore with
is tumadocd Uy the aun af deteaton win Ch sri ange Tot
pled (Hevlet-Pakerd nplfers adel 62164, 481 Nis Oslo Aves Tacos,
Calor) and recorded on the sp chat recorder ewe Packard Medel No. 080,130 Paricts Le ANDRES ET Al
‘Table 1, Pubwonry Function Testing
erm 20
Forced vial capacity (best of two a) Callin survey sprometer™
Masini vantary entation
waren B Goliog tne, Baites, Many
(TONE). Used to test over 250 patients with ALS, th E has heen
utilized in several drug trials.» ®* Our tost protocol (Tables 1 through
4) evolved Irom several years ol experience in collecting and analyzing data
to assure reliability and validity. As part of this process, many items have
been dropped, added, or modified over the past decade,
‘An important principle in designing a test batery is that the protocol
must address specific research questions. The TONE was developed to
measure the rate of disease progression in ALS and to determine how
much change occurs in a given time period. Test items that produce interval
data were required to allow adequate statistical analysis. If one is more
{interested in funetional outeomes, tests that produce ordinal data (MMT,
grading scales) may be appropriate, Analysis of ordinal data, however, can
only indicate general improvement or decline in performance. The following
summarizes our efforts to establish reliability and validity of the TONE.
Reliability
Reliability refers to the consistency or reproducibility of a measure-
ment.® Thus, when all variables are held constant, it is assumed that
change reliably measured from one trial to the next reflects true change
in the disease state. Test-retest variation is a function not only of testing or
measurement error but also human performance variation.* Positioning,
instructions to the patient, verbal encouragement, muscle contraction
relaxation time, and stabilization can influence the results.» Such variables
‘ean be somewhat controlled during isometrie strength testing when strict
standardized procedures are followed. However, despite minimizing testing
terror, it may be impossible to limit human performance variation to less
than 5 per cent.
‘Table 2 Timed Motor Actiitios
esr ee ure NEEDED once
ip ate “Stopwatch
Speech ster“ Stopwatch
one ding Singh dil Completing the ligt
Tht ant lane, stopwatch ‘umber 164.7172
Let han
Pen tet Purdue pegboard. uber of pgs pla in
Tht han “epateh ‘i soond
at har
Timed ot walk State Grossing the 1 ot maker
(Patent seated ia 38.6 an high lay a Font of Heh ale)
‘Model 3300, J-A Piston Corp, New Yor, New YorkQuavrrrarive Assesses oF Netnowuscutan Destcr ALS 131
‘Table 3. Newonuscular Functional Performance Seale
Speed Trott patent oak
nner etrmty Plies «8:5 om cube wong
Oa kg in 12S cm square box
Lower extremity Interview ard observe patent
smbolte
7
Normal
Sly showed, some suring of
Evidence of dysartri, sowed rte,
ely nerd
(Geveraly dye, may have
‘eto tli
Dyna el Yo wnderstand,
Dut sme pias undersea
Dysarts aly 2 few wow
coder
= Une, sooveral communication
seve bg wed
a en mal
{Gan Somplt, but wth
‘omen tecgiee
‘Canta bt canna complete
Unable
Unlined community ambultor
sina gat ater
Independent community ambulstr
‘ih or witout device, detectable
‘tt devious
[dependent howcholdambultos
‘wth or thot devices, ma ed
Sisto rar limited omy
“nuit with Supervision
‘sted ambulation the homes
tated trates
Sted fr 3 ted
‘Nina sist transfersParmicis L, ANDRES EF AL.
ie
‘gs os “THE-266 29" SmaINON sD NI jn u| WouIsOSs ZU! oMTMHUEN TET MEL “A
Tracy SopaUY so "og Seog DE
ow 2003
os 8908
06 w 0m
sup 2009
omg
open dy
xd serene soar
soyepunet spur amas 0518 mC mata
eras ip sang weeny 0 sony os w mpnes owing
a ensop 0 eand Dorspiets eng epg res
any eg oaosaa wm Bay Beapen, Pa104 EO fo POHPTF AMPE,‘Quanrrixtive AssessuNT oF NeuHOMUSCULAK Desicrt wy ALS 133
High reliability is often incorrectly assumed if dere is high test-retest
correlation. Te should be noted that a correlation coefficient of 1.00 des
‘not mean thatthe values of the Krst and second tals are identical. Rather
itimeans that they are pefetly matched. That i all first and second values
Ihave the same relative relation to each other. To. more fully assess
reproducility, It is. preferable to use per cent change of intraclass
Carrelaton coelficient (IOC) with interval data and per cent agreement with
ordinal data
To studying the testretst reliability of the TONE, 10 patients with
ALS and 35 normal subjets were ested twice within ato Shou period
Data were analyzed using both correlations and per cent variation, a shown
in Tables 5 through 8. Intrarater test-retest correlations ranged between t
= 0181 and 0.98 (mean absolute variation = 6.4. per cent) for normal
subjects; and = 0.92 and 0.99 (moan absolut variation = 8.9 per cent)
for ALS patients, Intaiter testretest corelations ranged between ¥ =
10.70 and 0.99 (mean absolute variation = 7 per cent) for normal subjects
and ¥ = 0,92 and 0.96 (mean absolute varation 7.4 per cent) for ALS
patients
“These results illustrate that different information can be obtained from
the same data using two different analyses. Though correlations were higher
for ALS patients, the normal contiols showed lest per cent testretest
variation. ‘This is most likely because the normal subjects were a more
hhomogencous group, whereas the patients with ALS had a wider spread in
their performance. Therefore, though the patients individually changed
from test to retest moro than the normal subjects, they more closely
‘maintained their relative postion compared with normal subjects who were
Lightly clustered. In adtition, our resulls demonstrated no dliflerence
Ipetween intra and interrater reliability. indicating high consistency tn the
testing procedure. This sguests that test-retest variation i due to subjeet
performance variation, rather than ervor in the testing method
Many investigators have demonstrated similar reliability using strain
gauges." 2 * Wiles and Kami tested knee extensor strength in 20
Patients with neuronvuseular diseases and found a test-retest correlation of
‘Table 5. Analyse of 3 to Sshour Tost-Retst for Toned Actvtios and Pubmonary
Punction: Intrarater Testing
TT etenee- «ne =
“tal 036 9 a1
Phone R ba os Sa
Phone L 95 om a
Pegard R oss on 4
goad ost om ike
ve bs O99 a4
vy bs om 103
er
aie
(Prom Anvires PL, Hdl We Fini Ly ot ok Qumaittive ator ssement
anyotrophi lateral leon Neurology 3457-98, 1968 wth persion)1m Parmicas Le ANDRES ET AL
‘Table 6. Analysis of 9- to S-Hour Test-Retst for Isometrie Force:
Intrarter Testing
NORA CoNTON T= 35) Taseanenn n= 1
Te ___Fercowthange pr __‘For ent cana
‘honk fx ny 3 om 32
Shaler ex on a om bo
Slams ext oo zo oo 5
Shney ext L 0. bo 30
Ell fer om 5a oo
Elbow fe 096 a3 bap
Eto ex 096 bs ba
Elbo et I Das 52 bar
Grp bas ta ba
Gap be bao
Hip ex R 095 Das
Hip ext bas her
ip flex oar h@
ip fee bas 9
Kee ee sr 4
Ke eat nes 4% 8
Ke fx O98 O98 46
Knee fee ase a9 so
Dorner ry on 40
Doras 1 Os 0s ur
i = Hexion
Ea = erenaim
(From Andres PL, Hedland W. Faison Ly etal: Quantiave motor asset in
‘unpotophic ater! sclerosis: Newoagy SBCTAIST- DA, 186 wth ermisin.)
able 7. Analysis of 3- to 5-Hour Test-Retest for Tuned Actielties and Pulmonary
Function: Interrater Testing
Toma conrwnis(e = 10) panne = 1)
Tem a Per cat hana a erent change
= os BI am oa
ta on ts os 03
Phone ® Om 26 3s a
Phone L 08 aT 000 a
Pega on 62 009 19
Pega L om 54 098 a1
re 004 so om %
ANY 03 a bop uu
er
Le ket
(rom Andves PL, Hedland W, inion L, eta: Quantitative motor asessment in
snnyoroic teal selves, Newtelogy SATN7-04, 1086 wth pean,Quavrrnynive Assrssweser or NeunoMUSCULAR DEFICIT IN ALS 135
“Table 8. Analysis of 3 to S-Hour Test Retest for Inmerie Fore:
Intersraer Testing
Tom connan f= 10) ras = 10)
Tew 7_Per cont change = __Perent chan
a
Shoulder er L095 3 om
Shoulder ek 098 33 098
Shoulder et LOST 52 090
Ew fe 8 099 ar aa
Ew fe ost Ba aa
Bl ex 98 7B 7
Elbo et 99 87 a8
np 098 58 om
Gal oor 6s om
ip ext 0.98 sa oot
Hip et Le ‘090 89 ba
Hip es hs er bas
ip es 1 ot wo ba
Knee ext it 9 ts a8
Knee ett aut 80 98
Knee fx oot 1 oa
Knee fs L ot 56 ost
Dean | REL 1 oar
Dovaiieron L_—O.85 Ba oar
= whe
Lake
fix = Henon
(from Andee PL, Hedland W, inion L, e sk Quunttate motor aueunent fa
sunyotophic trl sero, Neuro 367/897-81, 1988, with perio
= 0.57. With five examiners measuring 32 muscles, inter-rater coefficient
fof variation was 12.8 per cent. Scott et al. tested knee extensor strength
(Of 12 boys with DMD an observed a test-retest correlation coefficient over
1.5 hours ofr = 0.91
To be valid, a test must measure what it Is supposed to measure.
Because no method currently exists to measure motor neuron loss directly,
true validation of test methods to quantitate deficit in ALS is not possible.
Validity is best assessed by comparing the test protocol with a universally
accepted standard. Unfortunately, no such standard exists for quantitating
deficit in ALS. Therefore, validity must be assessed through theoretical
‘constructs
"The major component of the TQNE is measurement of voluntary
isometric force. Its theoretical framework assumes that a change in maximal
fsometric force is diredly related to motor neuron loss, Sobue et al.”
demonstrated a high correlation (r = 0.84) between the number of motor
neurons at autopsy and muscle strength tested 2 to 6 months before death.
Edwards," using myothermal measurements, has demonstrated in normal
subjects and patients that the quadriceps ean be made to contract as fully
jn a maximal voluntary isometric contraction as in a maximum stimulated136 Parwtets Le ANDNES Ev AL
‘5 negascores over tine for patient # 3
a) see MESA ea ee Tae
days singe disease onset
Figure 4, The decline ny ve fnctons of «sng patent over tne. The slopes appear
ears parallel
contraction. Therefore, measurement of maximal isometsic contraction is
dlinically the most direct means of ascessing the amount of functional muscle
‘We have included pulmonary function tests, diadochokinetc speech
rates, and timed functional activities in the TQNE, because these categories
provide adjunctive information about changes in functional capacity com-
monly seen in ALS. These tests are measures of the seqquclae of strength
loss and thus provide less direc evidence of disease progression. To validate
them, we charted decline in these functions against decline in isometric
strength.” A close correlation was found, and, therefore, the validity of
these tests was established using isometric strength loss as the standard
measure (Fig. 4).
‘Content validity addresses the minimal scope a measurement must
have in order to adequately reflect the variable of interest.” Originally, the
‘TONE consisted of more than 70 items and took 2 hours to complete. ‘To
improve efficiency while retaining comprchensiveness, a factor analysis was
performed to identify redundant items, Those that were closely clustered
by factor analysis were climinated. Items of low reliability. oF minimal
clinical use were also dropped. As shown in Tables 1 through 4, the test
protocol now consists of 25 quantitative items yielding interval data. that
measure motor performance and three graded items yielding ordinal level
data tha judge fonction abit; The test akes about 48 minutes to
complete,
SPECIFIC TEST CATEGORIES
Pulmonary Funetion
Progressive respiratory impairment is the primary cause of death in
ALS,"*and most ALS test protocols include some measurement of pulino-(Quasrrrarive AssessMenr oP NEUROMUSCULAR DEFICIT WN ALS 137
ry fanetion***** However, a decline in pulmonary function may be
influenced by factors such as emphysema, congestive heart failure, and
asthma, as well as by muscular weakness. We have therefore limited our
pulmonary studies to expratory volumes (sce Table 1), though others have
also tested inspiratory pressures." ® Test-retest correlations were high,
and variation between tests was generally under 10 per cent
Bulbar Function
We tested bulbar function (sce Table 2) by measuring speech rates in
the manner deseribed by Fletcher." This technique has been used previ-
ously in ALS drug trials**! Other investigators have also sed a descriptive
‘scale for chewing and swallowing *=® or videotape esophagrams.”” Wo
found that measuroment of speech rates provides a quick, reproducible,
and sensitive indicator of disease progression
‘Timed Functional Activities
‘Timed activities (see Table 2), such as phone dialing, the pegboard
test, and the timed alk, are inchided as assessments of upper motor
neuron (UMN) function, Direct clinical measurement of UMN loss is not
yet possible, but timed tests may provide indirect evidence, as spastic
manifest by decreased speed of movement and decreased ability to alternate
‘movement.® Efforts wero made to design timed test items to reduce the
‘effects of pure lower metor neuron (LMN) weakness. This was done by
‘sing, small loads and shert time intervals
Direct quantitative testing of spasticity continues to present « chal-
lenge, Quaiative sales have been used fo aes pst in ALS 2%
Norris graded stretch reflexes, plantar responses, and leg. “rigidity.”
Engel! focused on the patient's ability to perform bieyclestike movement,
the presence of conus, and the agility of walking. However, their methods
produce ordinal data and have not been validated.
Neuromuscular Functional Performance Seale
A functional scale was included in the TONE to provide a broad
description of a patient's functional status (see Table 3). However, these
data are not inchided in our statistical analysis
‘Maximal Isometric Force
Strain gauges produce accurate and valid interval data by which to
‘measure isometrie force (we Table 4)“ *! We have used the device for
this reason and also because the limb is positioned at 90 degrees, which
allows longitudinal studies despite contractures. In addition, testing at mic-
range serves to limit the effects of gravity and improve standardization.
Our strain gauge provides a reliable linear signal sensitive to 0.5 ke, and
able to tolerate up to 100 ky, Because the examiner is not required to
provide any resistance, the contraction force can be accurately measured
in patients with very weak or very strong musculature.138 Parmcts fe ANDRES ETAL
COMPOSITE SCORES (MEGASCORES)
Rationale
Individual data items were grouped into relevant categories. Scores
within each group were transformed and averaged into composite values of
“mezascores.” Data were grouped into megascores for several reasons.
First, the number of individval items in the test battery are too numerous
for elficient analyses. Rach patient can be assessed in 31 different ways at
‘each visit, and as longitudinal testing involves many patients, the amount
‘of data would be overwhelming. Second, the creation of composite scores
increases the reliability of measurement by balancing measurement errors,
By averaging several individual scores into one, the reliability of the
composite measure will be higher than any one component taken alone,
‘Third, composite scores reduce erroncons “significant” findings by reducing
the number of statistical comparisons. When many items are measured and
analyzed separately, the probability i high that one will produce a "signil=
jeant” effect by chance alone, Finally, we wanted to summarize function
by neurologic region
Statistical Methods
Grouping ems. A factor analysis was performed on data from a
tandom visit for each of the 176 patients with ALS in our data bank, to
determine the patterning of test-item intercorreations. Items regionally
‘and functionally related were also factor analytically related. This allowed
the creation of five megascores titled pulmonary function, bulbar function,
timed hand actcitos, isometric arm force, and tsometric leg force (Table
9)
Standardizing Scores. Individual raw scores were transformed to allow
equal weighting of raw values measured in diferent units or capacities
tsing2 score transformations (Table 10), In order toallow equal contribution
‘of an item with a relatively small capacity such as forced vital capacity
Table 9. Meunscore Compasttion
Hees 1 Palonary faction
HVC 1
Rd 1
Mega 2: Babar function
‘indecent: “po 1
Mega 2: Timed hand fnetion
Dialing a7-dat phone number 2
Purdue pexboard tet 2
Mega 4; Arm stent
loon sin gauge to measure ars meee strength 5
Jamar dynamometer 0 messre sop stent, 2
Megs 5: Lag sre
letoni tn gauge to measure log sometrie tenth 0
‘ToraL. Es‘Quawrtrarive Assesswtenr oF NeuHOMUSCULAR Davterr IW ALS 139
‘Table 10, Example of Z-Seore Transformation and Megascore Calculation
AS pom
Rawiacore AUS pops nes "DN __Eauation Zoo
ie oi 3s T “0.00
Mwy olen 6 -oar
Megencore Calelation
‘Mesa 1 (Pulmonary fanetion)
(FVC)—measured in single digits of liters—eompared with an item with a
relatively large capacity such as maximal voluntary ventilation (MVV)—
measured in double or triple digits of liters—the raw seores must be
transformed. The equation for z-score transformation is: z= raw score
‘minus the ALS population mean divided by the ALS population standard
deviation, Means and standard deviations were computed by averaging the
actual score for cach item on a random visit for each of the 176 patients in
the data bank. This procedure generated an ALS population mean and
standard deviation for each test item in which each patient had equal
weight.
‘Table 10 illustrates the use of z-seore transformations to convert FVC
and MVV to comparable units. Once z scores were calculated for each item
within a category (in this case FVC and MVV), the megascore was abtained
by averaging the z-score items. Megascores were only calculated when all
items in a group were available
Uses for ALS Megascores
We have used megiscores to analyze the natural history of ALS and
to test therapeutic efficary of drugs. Megascores of 50 patients with ALS
tested on eight or more occasions over 8 or more months were analyzed t0
determine deterioration rates and the course of the disease.*" A’ strong,
linearity of deterioration was seen, as shown in Figure 4
In a study to test the therapeutic efficacy of ehronie intrathecal
administration of TRH in ALS, we used megascores to calculate and
compare the slopes of deterioration before and during TRH therapy.”
SUMMARY
Several criteria must be met in developing a test battery for ALS.
First, the test items should be designed to answer the specif questions
being asked. If questiors involve the amount and rate of deterioration,
then the test items should generate interval data and show evidence that40 aravets L, ANDRUS ET AL
they reflect change in the disease itself. Second, the reliability of the test
items should be carefully determined. Are the variances due to testing
errors such as inconsistent positioning or changes in verbal instructions; oF
{do they reflect human perlormance variation? Third, the test items must
be sensitive to small changes, and one must be able to test the very strong
as well as the severely debilitated patient.
“The TONE was designed to measure degree of disease progression in
ALS. The major portion of the protocol consists of isometric force measure~
rent using an electronie strain gauge. Other quantitative segments include
tests of pulmonary function, speech rates, aud timed activities. Using
Score transformations, we have summarized these test items into five
‘megascores: pulmonary function, bulbar functon, timed hand activities,
frm strength, and leg strength, Meyascores have been used to characterize
the natural history of the disease and to judge therapentie offcacy
‘Quantitative methods to measure deficit in neuromuscular disease lag
far behind other technologie advances in medical research. Future advances
in the measurement of neuromuscular deficit depend on the commitment
of researchers to invest the time and resourves necessary to develop more
sensitive and reliable testing techniques.
REFERENCES
1. Andres PL, Hedland W, Faison Leal: Quantitative motor asesanent in ametropic
Tater sero: Newrabgy. 36707-0158,
2, Bensley WE Quaitatve te testing Paces and applications to research and
‘nial serves Arch Phys Mel Reha, 42098-45. 1961
2, Malley WO. Halt W, Caper Greta A double ind conte eal of bovine
‘aa gant romasi i ry ateral sclerosis, Nearby. S4:U70-1082,
19,
4. Broke MI, Gigs NC, Mendel J Clinic ris in Duchenne dy stop: The desi
rile pit: Mose Neve, 1185-197, 180
15. Rooke Ah Florence JM, Teller Sly etal Contolled tl of rotopin reusing
Thomone in smotropic neal cers. Newly, 36146-15108,
6 Cannio Jt, Caen Ja, Zavodnisk tae A double-blod,plscbo ontalled tril of
“TRI in amyotropbie aera selerot, Nourolgy, 36141-14, 1986,
1, Daniel ly Wortham C. Muscle Testing Techvave of Manual Hsamiation. Ea 4
lieiphin, WB. Saunders Co. 196.
5, defer Hoes i. Sterk JC. Rely of dysamomete i patents with newromuselar
‘ders, Eng Med 1160-174, 1082
9, kdands RHT" Hye 5. Methods of measuring muscle stengh fugve, Phystherspy
‘Ghsl-38, 107,
10. Edwards TT, Phsilogil nas of sheletal muscle weakness and tue. Cin St
ih0, 107%
11, Bagel WK, Signe T, Nelo: Het on weakness and spastiety in ALS of TRH,
Tancet, 222-70, 1085,
12, Fallt RJ Nos EH, Respiratory problems. In Mulder DW (ed The Diagnosis and
"Trestinnt of ALS Bonton, Houston Min Co, 1990.
1, Ftcher SC: Tune by-coant mearrement of dadocholanete syllable rt. J Spoech Hear
Ts, 15763-72017
14, Miso I, Peri J The okietieconopt of exer. Phys Ther, 47114, 1957
Is, Ide SA, scot OM, Gaurd CM The myometor: The development of las ol
Plysbtber @8 424-137, 1983.
16, Keleman J Hen W, Moray: Doug Ft Lain snot lective io amyotrophic
Titenl sleet Neurology, 32.315 316, 2952
17. Kendall H Kendal F, Wadsworth G: Muse Testing and Futon, Ka. 2 Balimoe,
‘Wiliams and Wilkin, 197,‘QuaNTEFATIVE ASSESSMENT OF NEUROMUSCULAR DivICrT IN ALS Mi
18. Kovlinger FN: Foundation of Behavioral Research, Ea, 2 New Yor, Halt, Rinehart and
“Winston 1973,
18, Mayhew TP, Rote JM: Measurement of muscle performance with instruments
Tt (ah sent a yl Ther, New Yor, Chul Lele
20, Medical Revearch Counce Aids to the taestiption of peripheral nerve injes. War
‘Memorandum (evted Ea) Lenton HMSO, 15, page 1-8
21, Mitsumota Tt Salgado ED, Nop eta Arye lara sevosts: Elects of
neve ttevenone and chrome sbeutenenws actin Oaropieoning
termone in eotrlled al, Newoogy 9-152-150,198
22, Molnar GE, Aletander, Le! N- Rely of quanta strength measurements in
shilden Auch Phys Med bab, 0218, 1979
25, Mins TT Heat W, Felmos M, Cooper, Gabba Cla les tothe ease af
'mjoeophieltralslevos Sera Ce): Newomsclar Disease. New York
ven Pres, 164
24, Manat TL. Bradley W: Amyotrophic later serosa Tyler Hl, Dawson D (ed
‘Current Neurlogy. Bost, Houghton Nii Co. 197,
25, Monat TL, Pasery CS, wy. etal Amantaie and guanine ae inet in
AES. Nevwoly. 8110511085, 195,
25, Minaat Te Andrey Pl, Boe, etal: The natural story of ALS. Ann Neural
815, 15.
21, Muna TH Tait J, Andres Fa a: neath fis of thyrottogtoseewsing hrane
(CH) in amyotrophic cal sclens A-contlled pt sty three pet
‘Rewrology, in pres 1
28, Norris FH: Canc PR Fall ota. Adinistiation of raniine tn amyotrophic
ater sclerosis. Newob. 24721-73817
25, Olnte MI Gerstein JC, Zbrekie Jota. Trane actor i ncfltve in smtrhie
Ite cern Ann Newel, 585-386, 1578,
90, Povin AB, Tourette WW, Syma K, el: Quantitative eth in anesent of
newrloge function, CRC Gi fey Bioeng 177-234, 18,
SL, Rivera VA, Graft, Deaton W, eal Madied sake venom fv ALS, Lack of ina
‘letiveness. Arch New, 7.61910, 180.
2. Hathstednf Mcanorement i Phyl Therapy: New York, Chorchl Livingstone, 1985.
‘SA. Sahrmann SA, Norton BJ The relatazaly of voluntary morereat to pati the
‘oper motor neuro syndrome. Ann Nedra, 3460-485, Desembes 107
54, Sool DM. Hyde SA. Goddard Ct al Quantitation of muscle Tint fier: A
wospecive stayin Duchenne msl dyateop. Miscle Nerve, 291-201, 182.
6, Sip, Nonparametric Sats forthe Behavioal Sciences, New York, MeCiew Hi
ck Co Bo.
96, Sobue G. Sash K, Talasashi A, et ak Degenerating compartment and fietioning
“compartment of motor nuroor in ALS Pole proces mor neuron kn
Neurology. 33854-657. BS
ST, Smidt GL Hers WM: Fars cotebuting othe regulation and lial astesnet of
muscular stent Phys Ther, 6 1285-1900, 152
38. Tandan I Bradley WG. Amvotonhic ister stern: Par 1. Clin fetes, tly,
Tan eli ses in mangement, Ang Neuro, 18271-28019,
99. vt der Plo 110, ‘Onstehuls HIGH, Rewelamp Jz Measuring muscle seat J
‘ew, 21-20-3051
40, VigosP, Watkins Me The fect of exercise muscular dystrophy, JAMA, 1974-48,
1966.
41, Wiles GM, Kuri ¥: The msorement of stent in ptt with perigheral neurone
ular digrdes J Newel Neurosurg Psychiatry, 1610061013. 1953.
42, Weight W. Muscle tiring nthe treatment of ifntle paral stow Med Surg J
48, ter FA Albop KG, Tyler PH: Asesmeat ofthe muscle sueagth is Duchenne muscu
trophy. Newogy, 47981-08107,
Newromnscilr Vat
Nev England Medial Center
171 Haron Avene, Box 373
Bostan, Masset B11]