schultz-analysis and measure lumbar loads bends and twists-1982

8/6/2019 Schultz-Analysis and Measure Lumbar Loads Bends and Twists-1982

1/7

ANALYSIS AND iMEASUREMENT OF LUMBAR TRUNKLOADS IN TASKS INVOLVING BENDS AND TWISTS

A. B. SCHULTZ G. B. J. ASDERSSOS*. K. H,+DERSPECK_R. ~RTESGRES.~ M. NORDIN* and R. BJ&K*Department of Materials Engineering. University of Illinois at Chicago Circle.

Box 4348 Chicago. IL 60650. U.S.A.

Abstract-Ten subjects performed isometric weight-holding and force-resisting work tasks whde standingupright. both with and without a twist of the trunk; with the trunk bent laterally; and in postures involvingcombinations of bending and twisting. The lumbar trunk muscle contraction forces and the lumbar spinecompression and shear forces imposed by these tasks were predicted using a biomechanical model.hlyoelectric activity was recorded quantitatively at eight locations over the back muscles and at fourlocations over the abdominal wall muscles. Correlation coefficients from 0.67 to 0.88 were found between thepredicted muscle contraction forces and the measured myoelectric activities when the predictions were madeso as to minimize muscle contraction force per unit area. Trunk t\\isting and lateral bending were found toload the spine and trunk muscles less than trunk tlexion or holding of weights in front of the body.

ISTRODCCTIONIn a previous study ;1 mathematical model to de-termine the loads imposed on the trunk muscles byphysical taks was validated using electromyography(Andersson rt nl., 1979). The validation was based onlaboratory measurements of the myoelectric activities ofseveral trunk muscles when healthy subjects wereasked to perform tasks in which their trunks wereupright. The model predictions of the muscle tensionswere highly correlated to the measured myoelectricactivities. In actual life, more complex postures areoften assumed ; postures involving twisting and lateralbending of the trunk. The aims of the present studywere to test the validity of the model in these morecomplex circumstances, to develop a better under-standing of the biomechanics of these activities. and todevelop more extensively practical methods to de-termine the loads imposed on the trunk by work taskperformance.

Ten male university students participated. Theirages ranged from 23 to 34, with mean of 28.5 yr;heightsrangedfrom 166 to 194, with a mean of 183cm;and weights ranged from 65 to 97. with a mean of75.5 kg. All subjects were in good health and none hada previous history of significant low-back or otherhealth impairment. Clinical examination did not re-veal any diseases or abnormalities.

*Department ofOrthopaedic Surgery I,Sahlgren Hospital.S-413 45 Gdteborg. Sweden.+Department of Clinical Neurophysiology, Sahlgren Hos-

pital, S-413 45 GGteborg. Sweden.

A total of 20 tasks were studied. All stud& wcrcperformed with the subjects standing in a refercnccframe with feet and legs in comfortable positions andwith their pelvises strapped to a support board. In thefirst seven of the tasks a harness was placed around thechest just under the arms. Cords were run horizontallyfrom the harness over sets of pulleys and weights werehung from the cords. The subjects were asked toremain upright and resist the weight applications. Inthe first of these tasks the subjects were standingrelaxed and no weight was applied. In flexion andextension resist tasks, a 15 kg weight was applied, eitherdirectly anterior or posterior to the trunk. In two rightlateral bending-resist tasks, weights of IO kg and then20 kg were applied directly to the right of the subjects.In two twist-resist tasks, txo IOkg lveights and thentwo 20 kg \vrights bvere applied to impose pure lonpitu--dinal twisting moments. To that purpose, one weight~vas hung directly in front of the left anterior axillqfold, and one directly behind the right posteriolasillary fold.

In the next three tasks, the subjectscontinucd to fact:forward. In one they extended both arms laterally andheld weights of 2 kg in each hand. In the other two,they bent their trunks to the right so that it w;1sinclined 20 laterally as measured by a goniometrrplaced at the T7 level of the back. They then placedtheir hands one at each shoulder at first holding nobveight. and then holding 4 kg in the right hand,

For the next eight tasks, the subjects twisted to theright so that the linejoining their two shoulders made a45 angle u-ith its original direction in facing forwardThe pelvis was held in its original position, and notallowed to rotate. In this configuration the first taskwas to stand relaxed. A second was to extend botharms latrrall~- while holding 2 kg in each hand. A third\vas to extend only the left arm. and a fourth to extend


2/7

6 7 0 A. B. SCHI.LTZ er d.only the right arm. both while holding 2 kg in theextended hand. In the other four tasks performed witha -l5- shoulder twist. both hands were held at the chestand then both hands were extended anteriorly. In eachposition of the hands, no weight was held at first andthen a 2 kg weight was held in each hand.In the final two tasks. the shoulders were againtwisted 15. The trunk was then flexed 30 at the hips,measured with a goniometer. In one task. the handswere held at the chest: in the second they wereextended anteriorly while holding 2 kg weights in each.

\IEASUREJIEST OF COSFICURATION DATA

Three targets vvere taped onto the subjects rightside ; one transverse to the mass center of the head, onetransverse to the mass center of the trunk at approx-imately the T9 level, and one transverse to the centerof the L3 vertebra. Scales were placed on the referenceframe in front of and on the right side of the subject.The horizontal scales, two plumb-bobs and the targetsvverc used to estimate both the antero-posterior andthe mcdio-lateral locations of the mass center of thehead. the mass center of the trunk, the center of L3, andthe center of the right and left hands.

In the horizontal force-resist tasks, the verticaldistances from the cord to the L3 level were measuredwith tape in the flexion-extension, and lateral bend-resist tasks. The horizontal distance between the twocords in the twist-resist task was also measured with atape. This was done so that the moments of theexternal forces at the L3 center could be computed.The width and depth of the trunk at the L3 level werealso measured in each subject.

MEASUREhlEST OF MYOELECTRIC ACTIVITY

Myoelectric signals were picked up by twelve pairsof bipolar, recessed surface electrodes. Six of them wereplaced on the right and six on the left side of the bodyso as to have bilateral symmetry in electrode pairlocations. The bipolar axis was aligned parallel to thedirection of the muscle fibers in each case. Posteriorly,electrode pairs were placed 3 cm lateral of the midlineof the spine at the C4, T8 and L3 levels. Anotherelectrode pair was placed 6cm lateral of the midline atthe L3 level. Anteriorly. a pair ofelectrodes was placedover the rectus abdominis muscle at the level of theumbilicus and 2cm lateral of the midline. Anotherpair was placed over the oblique abdominal muscles,3 cm medial and superior to the anterior iliac spines. Aground electrode was placed on the right ankle.

The myoelectic signals were fed to differentialpreamplifiers contained in a 0.4 kg box which wasstrapped to the subjectschest. The signals were furtheramplified and recorded on magnetic tape. For analysis,the tape signals were first fed to a detector and then tothe analog-to-digital converter ofa PDP-I5 computer.The true RMS detector included a low-pass filter witha 3 dB frequency limit set at 0.8 Hz. The analog-to-

digital conversion rate for the filtered Rh1.S signal was6 Hz.

All tasks were performed quasi-statically. Oncesteady-state conditions were achieved. the myoelectricrecording equipment was started and configurationmeasurements were made. Recordings were made for15 s, after which the subject rested for one minute. Foreach of the 20 tasks, the means over the 10 subjects ofthe myoelectric signal levels at each electrode locationwere computed. In addition a calculation was made ofthe mean of these means over the two L3 electrodepairs on each side (3cm and 6cm lateral to themidline).

AMLYSIS dF TRUNK MUSCLE COSTRACTION FORCFSAND SPINE COMPRESSI OS LOADS

The analysis of the trunk internal forces imposed byeach task on each subject was carried out in two mainsteps. The first was to compute the net reaction at theL3 transverse trunk cross-section, and the second toestimate a set of trunk muscle contraction forces andmotion segment compression and shear loads acting atL3 that could provide that net reaction.

The net reaction was computed from equilibriumconsiderations. All external and body segment weightloads and their lines of action were substantiallyknown from the configuration measurements taken.The three mutually perpendicular components of thenet reaction force and the corresponding three com-ponents of the net reaction moment that must besupplied to equilibrate the trunk segment above thatcross-section were determined from the six availableequations of equilibrium.

Toestimate the muscleand motion segment internalforces, the L3 motion segment was assumed to providecompression and shear resistances, but no significantmoment resistance. Intra-abdominal pressure wasassumed to be zero. Five bilateral single muscleequivalents were assumed to act across the L3 section(Fig. 1). These represented the rectus abdominis. theinternal and external oblique abdominal muscles. theerector spinae muscles, and those parts of the latis-simus dorsi muscles which are present in the lumbarregion. Linear programming was used to solve forthese ten unknown muscle forces so that (1) the threeequations of moment equilibrium were satisfied; (2)the muscles experienced only tensile forces; (3) theprescribed value of maximum contraction intensitywas not exceeded and (4) an objective function wasgiven minimum value.

Each of the two objective functions was used in turn.One led to the minimization of the compression on theL3 motion segment. When this was used, the maxi-mum contraction intensity was limited to lOON/cmin any muscle. The second led to the approximateminimization of the largest muscle contraction in-tensity. To do this, the maximum allowed contraction


3/7

Analysis and measurement of lumbar trunk 671

Inten sity ga s first set to 5 S c m . If no solutio n to thelinea r p rogram ming prob lem existed a t this intensity.the allowe d intensit! was increa sed by 5N cm a ndan othe r solution trial wa s ma de . This intensity in-c rsme nta tion proc ess wa s repea ted until a solution~~a5 ound.

traction force estimate. The m>oelcctric acttvlty in theob lique a bd om inal electrodes MS correlated on ea chside with the sum of the internal and extsrnal nb liqu cco ntraction force estimates. Eac h resrrsslon tnvohed10 da ta pa irs. one for ea c h task.

Onc e the linea r program ming problem wa s sohrd.the rema ining eq uations of force eq uilibrium !\ eresolLed for the two mo tion seg me nt shea r force s an d forthe co mp ression if that had not a lread y b een found. Bqthese proc ed ures, estima tes of all of the c ont rac tionforces in the ten muscle eq uivalents a nd the threemo tion seg me nt resista nc e force s we re ca lculate d. Theproce dures a re de scribed in grea ter det ail by Schultzand And ersson (1981 ). Trunk c ross-sec tiona l peo -met ry was sca led for e ac h subjec t from his measuredtrunk diameters in prop ortions spe c ified in Fig. I.

HESCLlS

Forws irrlprwd OPI r/lr L? r,roriful .wc/~mrfrThe me ans ove r the ten subjec ts of the co mp ression

loads imp osed on the L3 mo tion segm ent by the 20ta sks ranp c d from 470 to 9610s (Table I I. The meanan terop ostc rior or lateral shea r forces imp osed o n 1~3we re at most 2OON. The mo tion-seg me nt lo: lds calcu-lated when the muscle contraction intensity \ \ asminimized did not dilrer m uch from those calculatedwhen the spine c omp ression \ \ as minim&d.

The mean over the 10 subject s ofea ch force estimatein ea c h of the 20 tasks w as c alc ulated . Linear re-grcssion analvses \ \ crt then ma de betwee n the forcecstima tc mea ns and the myoe lectric signal lcvcl mea ns.The myoe lectric ac tivity p icked up by the rcc tusahd om inis electrodes and the sum of the ac tivitiespicked up b y the lumba r erecto r spinae elec trod es wereco rrelate d on ea ch side with the co rrespo nding co n-

The mea ns over the ten subjec ts of the ca lculate dmuscle contraction forces on OIIC side in the erec torspina c musc les range d to ap proximat ely 1 100 N. w dthose in the ab d om inal ob lique s to 290 N (Tab le 2). Inthe rcctus ab do minis. the ca lculate d force per side wasa t mo st 17ON. and for mo st of the ta sks. it wa s zero.The choice of the objcc tivc function affected theco ntraction force s ca lculate d, but only m ode rately.Clear distinctions could be mad e bet\ \ esn thoseexercises which were more a nd those which \ vere Icssstrenuous, rega rdless of which of the t\ \ o ob jcc tivcfunc tions wa s used . In the minim um intensity WI-utions, the largest req uired mean muscle contractionintensity wa s 44N:c m (440 kPa).

Mo st of the a c tivities stud ied im po sed only sma llload s on the spine an d c alled for only sma ll musclec ont rac tion forces. The e xerc ises involving tlvisted orlaterally be nt bod y co nfigurations did not load thetrunk structures sign ific a ntly mo re tha n sagitally-symmetric configured exercises involving comparableexternal load s. The largest spine load s. and gene rallythe largest muscle c ontrac tion forces, were imposed bythe tw o exerc ises invo lving 30 of trunk tlexion. There.even when the hands we re plac ed close to the chest andno we ights we re held, the loads impo sed ~rell exceed edthose imp osed by an): of the othe r 18 exercises.

Fig. I. Schematic of the model used for internal forceestimation. The fi\e bilateral pairs of single muscle equiva-lents represent the rectus abdominis, the internal and externaloblique abdominal. the erector spinae, and those parts of thelatissimus dorsi muscles which cut the trunk sectioningplant. Contraction forces in these model muscles are denotedR. I. .V. E and L respectively. with the subscripts denoting leftand right sides. Inclination an$es. j?. 6 and ; were all xt to45 Motion segment compresslon force is denoied C, ante-rior shear force S, and right lateral shear force S,. Abdominalcavity pressure resultant P here \\a set to zero. Muscle cross-sectional area per side were taken as. respectively for R, 1. .y Eand L: 0.006. 0.0168. 0.0148, 0.0389 and 0.0037 times theproduct of trunk cross section depth and width. C entroidalofTsets in the anteroposterior direction were. in the sameorder. taken as 0.54. 0.19. 0.19, 0.X and 0.28 times trunkdepth. In the lateral direction. they were taken as 0.12. 0.45.

0.45. 0.18 and 0.171 times trunk width.

The me an ca lculated muscle contrac tion force s andthe mea n mea sured myoeloc tric ac tivities were inge ne ral we ll-co rrela ted (Table 3. Figs 2-4). With theminimum co ntraction intensity assump tion. the linea rc orrelat ion rang ed from 0.67 to 0.88 ove r the six mu sc leforce/ ac tivity co rrelations ma de . The c oefficientswere mo re variab le when the minimum co mp ressionassum pt ion wa s used , rang ing from 0.34 to 0.92. Thepoo rest force-ac tivity co rrelations were ob tained forthe ab do minal ob lique muscles when the minimumco mp ression ob jective function was used to predictthe muscle tenslons.


4/7

A. B. SCHLLTZet al.Table I. Mean predicted forces (N) on the L3 motion segment using the two different objective functions

Task performed Minimum compression solution Slinimum intensity solutionCompr. Latl shear AP shear Compr. Latl shear AP shearStumlitlqwithout nistStand relaxed 470 0 0 5Oil 0 0Resist Hex. 15 kg* 1390 0 150 1160 0 150exm. 15kg 690 0 - 150 970 0 - 170latl bend. IOkg 620 90 40 670 60 30IaU bend, 2Okg 850 150 50 1150 110 -10twist, 10 kg 730 150 100 860 30 40twist. 20 kg 970 160 130 1100 50 -50Bend right, hold 4 kg 840 110 40 940 10 0Arms latl, hold 2 + 2 kg 500 0 0 520 0 0

Sttrrldi~ly rc.itll 45 faistStand relaxed 480 0 0 500 0 10Arms latl. hold Z + Z kg 640 0 0 680 0 0Hands at chest, 2 + 2 kg 780 20 10 SlO 0 0Arms anterior, 2 + 2 kg 1190 40 20 1350 10 0SrrrrGry twisred trrd flexed 30Hands at chest 1700 0 180 1800 0 180Arms anterior, 2 + 2 kg 2290 0 200 2610 0 200

*The mean moments imposed on L3 by these weights in the six resist tasks were 44.3, 40.3, 27.6, 55.8, 25.6 and 51.2N m.

Tabulations of the results obtained that providemore detail than that practical to give here areavailable from the authors upon request.

DISCUSSIOS

In our earlier studies of the performance of simplephysical tasks, WC found that our model-based pre-dictions of trunk internal forces uniformly correlatedwell with our experimental measurements (Andersson,et ~1.. 1979; Schultz and Andersson, 1981 ; Schultz, et

(II., 1982a ; Schultz, et (II.. 1982b). In the present study,the regression coefficients are less impressive. But, theweaknesses in the correlations are not qualitative;there are no major discrepancies in the force/activityrelationships. When the model predicted that a muscleshould be contracting strongly, that muscle con-sistently showed a comparatively large myoelectricactivity. When the predicted contraction was small, themyoelectric activity was consistently small. The modelseems to be valid in at least a semi-quantitative sense,even for the complex maneuvers studied here.

Two sources of quantitative shortcomings in the

Table 2. Mean trunk muscle contraction forces (N) required for some of the tasks. These \vere calculated using the minimumintensity solutions. Rectus abdominis and latissimus dorsi contraction forces were at most 90 and 110N. E = erector spinae, I= internal abdominal oblique, R = external abdominal oblique. Subscripts indicate left or right sideTask performed E, ER 1, 1, x, xl7 Required intensity (N/cm)

Stmthg without twistStand relaxed 60 60 0 0 0 0 6Resist flex. 15 kg 510 510 0 0 0 0 23extn, 15 kg 0 0 160 160 140 140 23right bend 1Okg 110 40 60 20 70 20 14right bend 20 kg 310 120 50 50 120 40 23t&t, 10 kg 60 210 170 0 0 110 19twist, 20 kg 80 260 290 0 d 220 31Bend right. hold 4 kg 340 30 10 10 90 0 15Arms latl., hold 2 + 2 kg 50 50 0 0 0 0 6Sttrmhy with 45 mistStand relaxed 50 50 0 0 10 10 7Arms latl.. hold

2+2kg 100 150 0 10 0 10 9Hands at chest, 2 + 2 kg 250 140 10 0 10 0 11Arms anterior. 2 + 2 kg 580 210 90 0 90 0 24Srtrt~cl wist& trfrtl.&Wl 30Hands at chest 700 670 10 0 10 0 30Arms ant., 2 + 2 kg 1120 860 110 0 110 0 44


5/7

Analysis and measurement of lumbar trunk

TJble 3. Correlation coefficients betueen mean calculated muscle contraction forcerand mean measured myoelectric activities

MuscleErector spmae

Left sideRight side

Rcctus abdominisLeft sideRight side

Sum of internal andexternal abdominaloblique forces

Left sideRight side

Function used to predict muscle forceshlinimum 1lmlmum

spine compression conxicnon intsns~ty

0.86 0.850.60 U.710.82 0.700.92 0.83

0.55 0.770.34 0.67

0

5 lOOO-z /fl /c /$k 0* /zijz 500- /o/2 /5 O /o.s 0 0ii Left Srde

0 I I0 50 100

Mea Myoelectrlc Signalevel (Y)Fig. 2. Predicted contraction forces vs measured myoelectricactivities in back muscles on the left side. Each point showsten-subject mean values of forces and ofelectrical activities forone task. The regression line for the data pairs is shoivn. Datafor tasks calling for zero muscle tensions are not sho%n, butHere included in the linear regression analysis. The forcesshown are those predicted using the minimum contraction

Fig. 3. Predicted contraction forces vs measured m!oelectricactivities in back muscles on the right side. See Fig. 2 caption

for explanation.

intensity assumption.

correlations need to be considered; the surfacemyoelectric activities might not relate to the con-traction forces in specific muscles accurately enough,and/or the force predictions might lack quantitativeprecision. The former source is particularly trouble-some in the oblique abdominal muscles, because theresults show clearly that those electrodes picked up theactivity in both the external and internal abdominaloblique muscles. Since we do not yet know fully how tointerpret those signals, we arbitrarily correlated themwith the sum of the contraction forces in the twomuscles. More research is needed to learn how tobetter interpret those signals. Activity measurementsutilizing electrodes implanted into each of those twomuscles might contribute to that goal.

The accuracy of the force predictionsdepends on the

task being analyzed. Predictions of contraction forcesin tasks involving extension, lateral bending andtwisting moments are more questionable than thoseinvolving flexion moments. Tasks involving Rexionmoments call primarily on the erector spinae muscles;tasks involving extension, lateral bending and twistingmoments tend to recruit many of the lumbar trunkmuscles. The fibers of the erector spinae all lie close tothe centroid of the muscle cross-sectional area, so thatthese muscles have homogeneous and easily definedlines of action. The fibers of the latissimus dorsi andthe two abdominal oblique muscles are spread out.These muscles have more variable and less easilydefined lines of action. The erector spinae with theilalmost purely longitudinal lines of action have rela-tively simple mechanical actions. while the latissimusdorsi and the oblique abdominal muscles with theircurved and inclined lines of action have more diverseactions. So that is easier to represent accurately theerector spinae muscles in a model by single muscleequivalents than it is to represent the latissimus dorsior the oblique abdominal muscles.

0

5ooc

0/

/

/ 0/

s 0 $9 I I0 50 100Mean Myoelectr~c Sqnal Level (pV)


6/7

3z8 500 Right Sde5f 0

Mean Myoelectnc Signal Level (pV)Fg. 4. Pred icted co ntractio n forces vs mea sured myo elec tricac tivities in the abd om inal obliq ue muscles. See Fig. 2 captionfor explanation. The tensions indicated are the algebraic sumsof the predicted co ntrac tion forces in the internal and external

abdominal oblique muscles.

Moreover, little is known about how muscles arerecruited when several options of the accomplishmentof a task are available. Muscle recruitment patternsunder those circumstances probably are not unique;they can vary from person to person and from time totime in any one person. Tasks involving primarilyAexion moments are not of this kind. The erectorspinae are the major contributors to resisting trunkRexion, so essentially only one option is available formost Rexion resistance tasks. Several options areavailable to provide each of extension, lateral bendingand twist resistances, recruiting different combinationsof the rectus abdominus, the oblique abdominals, theerector spinae and the latissimus dorsi. For thesereasons, the adequacy of the muscle representationand the choice of the objective function are seldomcritical in the analysis of tasks that involve primarilyRexion moments; they are much more critical in theanalysis of tasks that involve extension, lateral bendingand twisting moments. Our earlier studies consideredprimarily tasks of the former kind ; the present studyconsiders a number of tasks of the latter kind.

During the most strenuous of the perfomancesstudied some subjects showed or expressed slightfatigue. However. every subject was able to performevery task for the 15 s period required and the next taskwas begun only after the subject said he was rested.There were no indications that fatigue affectedmyoelectric signal quality. Intra-abdominal cavitypressure was assumed to be zero in biomechanicalanalyses based on our previous experience. Schultz etrtl. (1982b) found that in I5 squasi-static performancesof tasks similar to those studied here, intra-abdominalpressures were rather small (1-6 kPa) and provided arelief of spine compression averaging only 14%. Sincein these present experiments we did not measure intra-

abdominal pressure, it seemed reasonable to make thisassumption about it.

Similarly, in earlier work we examined the use ofmyoelectric activity data normalized to the maximumelectric activity recorded by that electrode for thatsubject. This normalizarion procedure did not signi-ficantly reduce the interindividual variation. For thisreason ue used myoelectric activity data in absoluteform for this study. Myoelectric activity in the lumbarportions of the latissimus dorsi was not recorded, so itis not known how good our model was in predictingtheir contraction forces.

It is sometimes said that performances of tasksinvolving trunk twisting or lateral bending. or com-binations of these with anterior bending. unduly loadthe lumbar trunk structures. Neither the model ana-lyses nor the myoelectric measurements indicate thatthis is the case. Whenever two similar tasks wereperformed. one with and one without a trunk twist, thepredicted loads or the measured myoelectric activitieswere found to differ only slightly. Our results suggestthat twisting does not have much effect on trunk loadsin the situations we studied.

The need to support weights placed \vell anterior tothe spine, as when weights are held by anteriorlyoutstretched hands or when the trunk is flexed so thatupper body segment mass centers move out anteriorly,loads the spine structures and trunk muscles heavily.This occurs because the flexion moments these weightsproduce must be balanced by contractions of theposterior back muscles. Since these muscles act on thespine through short moment arms, they must contractstrongly to achieve the desired moment balance. Butwhether or not such tasks involve a trunk twist seemsto make little difference. It is the Aexion moment ratherthan the twist that is responsible for the large loads.For example, in the present experiments, holding 4 kgin the anterior outstretched hands with the trunkflexed 30 and twisted 45 imposed an L3 compressionforce of at least 2290N and required total posteriorback muscle longitudinal contraction forces of at least1950N. Schultz, er nl. (1982a) reported that a similaractivity, when performed without the trunk twist.imposed an L3 compression of 2220 N and required amuscle contraction force of approximately 1870 N.The corresponding figures for holding 4 kg in thenearly outstretched arms while standing upright with-out a twist were 960N and 550N while the same taskperformed with fully outstretched arms and with a 45degree trunk twist called for forces of I l90N and750N. In determining the loads imposed on the spine,trunk forward bending is important while trunktwisting seems to matter little.

Similarly, none of the studied activities involving alateral bend of the trunk or the lateral holding ofweights loaded the spine or the trunk muscles heavily.The imposition of lateral bending moments can loadthe spine moderately. but not nearly as much as theimposition of flexion moments can. This occurs be-cause the trunk cannot be laterally offset very much,


7/7

Analysis and measurementof lumbar trunk 675

so that the mo me nt imp osed by the laterally offsettrunk weight ca nnot bec ome \ ery large. In ad dition.the lateral a bd om inal wa ll muscles ac t on the spinethrough a relatively large mo ment arm, so they neednot contrac t strongly to co unterbalanc e the offset\ \ e igh t moment .

The bio mc c ha nica l mod el used in these stud ies toca lculate the loads imp osed on the lumba r trunkstructures d uring pe rforma nc e of ta sks invohfing t runkbe nding and twisting predicts those load s mod eratelywell. but not as well as it ca n predic t the load s imp osedby sag ittally symm etric ta sks that tend only to flex thetrunk.

Further resea rch is nee de d to exam ine pa tterns usedfor the recruitme nt of the trunk muscles whe n direrentop tions a rc a va ilab le for tha t recruitmen t. Trunkmt ~ s c i c myoe lcct ric ac tivities need to be mea sured withintram uscular electrodes, to de linea te mo re ca refully\ bhat muscles contrac t. to what deg ree. and to whatextent the different pa rts of a muscle c an co ntractindc pend cntly of each other.

Trunk tw isting b) itself do es not see m to loa d thespine or trunk muscles very much. Trunk Hesion andother activities that impose largs Aexion moments loadthese trunk structures hea bil!. whet her or not ac c nm-panied by twistinp.

.-l ,l\ n~,i~lrtlr/ ~~~~~,~~rs-The support of the Swedish Iiork En-vironment Fund, the S\*edish \led~caI Research Council andUS Public Health Senice Grants NS 151(x) and OH Cw)jlJand Development Atlard AlI OOO~9 1s gratefullyackno\r Iedged.

Anderson, G. B. J., Cirtengren, R. and Schultz. .A. B. (19791Analysis and measurement of the loudson the lumbar spineduring work at a table. J. Birww~hmics 13. 5 13-520.

Schultz, A. B. and Andersson, G. B. J. (19Sl) .Analysis Iofloads on the lumbar spine. Spirlr 6. 76-81.

Schultz. A., Andersson, G., Ortengren, R.. Bjork. R. andNordm. M. (1983) Analysis and quantitative myoelectricmeasurements of loads on the lumbar spine \\hen holdingweights in standing postures. Spry (in press,.

Sc hultz. A., And ersson , G., Ortrnp ren. R., Haderspeck. K.and Nachemson. A. (1982b). Loads on the lumbar spine:validation of a biomechunicul analysis by measurements ofintradiscal pressures and myoelcctric signalz J. &~r ,I/SW!/. 64A, 7 13~731.

schultz-analysis and measure lumbar loads bends and twists-1982

Documents