UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of valgus bracing on the 3D kinematics of gait in patients with osteoarthritis of the knee Davidson, Peter Lawrence 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-0382.pdf [ 1.69MB ]
JSON: 831-1.0077138.json
JSON-LD: 831-1.0077138-ld.json
RDF/XML (Pretty): 831-1.0077138-rdf.xml
RDF/JSON: 831-1.0077138-rdf.json
Turtle: 831-1.0077138-turtle.txt
N-Triples: 831-1.0077138-rdf-ntriples.txt
Original Record: 831-1.0077138-source.json
Full Text

Full Text

THE EFFECTS OF VALGUS BRACiNG ON THE3D KINEMATICS OF GAIT IN PATIENTSWITH OSTEOARTHRTTIS OF THE KNEE.byPETER LAWRENCE DAVIDSONB.A.Sc. University ofBritish Columbia, 1987A THESIS SUBM[TTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(School ofHuman Kinetics)We accept this thesis as conformingto the required standardTHE UMVERSITY OF BRITISH COLUMBIAAugust 1994©Peter Lawrence Davidson, 1994in presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________________________Department of //l4avi k’I-?c//EsThe University of British ColumbiaVancouver, CanadaDate A3 / //DE-6 (2188)11ABSTRACTClinical evidence suggests that there can be a reduction of pain by wearing a valgusknee brace for treatment of unicompartmental medial osteoarthritis of the knee (Horlick andLoomer, 1991). It has been suggested that the brace alters the orientation of the thigh andshank segments during gait in such a manner as to reduce the inter-joint pressure of theafflicted compartment. This study was designed to test the hypothesis that the knee bracealters the alignment of the shank segment with respect to the thigh segment.As a part of this study twelve subjects was recruited. Each had been diagnosed withmedial compartment osteoarthritis of one knee and for treatment, prescribed a valgus kneebrace (Generation II Orthotics Unloader). Prior to the study, the subjects had worn thebrace for over one month on a daily basis and report pain relief compared to the time beforethey received their brace.While walking on a treadmill at either of two speeds the each of the subjects was filmedby two gen-locked video cameras of 60 Hz. A total of five reflective markers were placedon the thigh and shank segments of the afflicted leg. After digitizing the marker pointsrecorded by two cameras, the three dimensional coordinates were determined using thedirect linear transformation method. From these data the movements and flexion, varus andaxial rotation angles of the thigh and shank segments in a global coordinate system weredetermined. The data analyzed was the normalized average of five sequential step cycleswhere each cycle was the time from one heel-strike to the next heel-strike of the same footof the afflicted leg. The values of the angles were calibrated with respect to the standingposition,”no-brace” condition, i.e. those readings were taken to be the reference point foreach subject.The study was a repeated measured design with two factors and 23 dependent variables.The factors of interest were brace condition: i.e., no-brace and brace, and walking speed:i.e., medium and fast. Thus, there were four conditions defined by combining the levels of111the two factors. Each of the 12 subjects were tested at every condition. Therefore, thestudy was a (12:subject X 2:brace X 2:speed) RM design. The statistical analysisconsisted of twenty three ANOVAs at the p<O.05 significance level. The hypotheses werethat there there was no difference of the percent cycle timing during the stance phase andno difference of the flexion angle between the no-brace and brace conditions across allsubjects at either walking speed. But it was expected that there will be a decrease in thecoronal angle (i.e., towards valgus) and of the axial angle (i.e., internal rotation) of thethigh and shank and this difference will be greater at the fast speed compared to the mediumspeed, i.e., there is an interaction effect. All the hypotheses were tested at four specificpoints in time during the stance phase.The results showed that the brace did have a significant effect but not in the wayanticipated. The brace had no effect on the thigh coronal angle but was significant inreducing the shank coronal angle into varus at toe-off The brace also prevented fullextension during mid-stance but this was believed to be caused by resisting valgus forces inthe coronal plane transmitted through the braces helical strap. This same strap is alsobelieved to be responsible for axial forces that created the significant external rotation ofthe thigh axial angle throughout the stance phase and the external rotation of the shankaxial angle during knee flexion. Dynamic forces of the brace hinge may explain theunexpected internal rotation of the shank axial angle that accompanied knee extension.There was a possibility that the subject interacted with the brace in a complex way toproduce some of the observed motion.ivTABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTS ivLISTOFFIGURES viLIST OF TABLES viiiNTRODUCTION 1Statement of the Problem 4Hypothesis 5Delimitations 6Definitions ofTerms 7LITERATURE REVIEW 9METHODOLOGY 14Subject Selection 14Equipment 14Goniometer System 14Video and Treadmill System 15Subject Setup 17Data Collection 18Data Analysis 19RESULTS 21Flexion Angle 21Thigh Coronal Angle 23Shank Coronal Angle 24Thigh Axial Angle 26Shank Axial Angle 28Event Times 31Results Summary 32DISCUSSION 33Flexion Angle and Shank Coronal Angle 34VThigh Axial Angle and Shank Axial Angle 36Subject Interaction 38FUTURE RESEARCH 42CONCLUSION 45REFERENCES 46APPENDIX A: SUBJECT RECRUITMENT FORM AND SUBJECTCONSENT FORM 50APPENDIX B : THIGH AXIAL ANGLE AND SHANK AXIAL ANGLEVECTOR FORMULAS 55APPENDIX C: ANALYSIS OF VARIANCE DATA 59APPENDIX D : iNDIVIDUAL SUBJECT AND GROUP AVERAGEANGLE GRAPHS 72vLIST OF FIGURESFIGURE 1: CORONAL ANGLE 8FIGURE 2: VALGUS KNEE BRACE 8FIGURE 3: NORMAL KNEE 10FIGURE 4: VARUS KNEE 10FIGURE 5: KNEE FORCES BEFORE BRACE 12FIGURE 6: BRACE AN]) KNEE FORCES 12FIGURE 7: EQUIPMENT SETUP 16FIGURE 8: NO-BRACE CONDITION 16FIGURE 9: BRACE CONDITION 16FIGURE 10: MARKER AND AXES POSITIONS 17FIGURE 11: EVENT LOCATIONS ON A TYPICAL FLEXIONANGLE CURVE 19FIGURE 12: GROUP AVERAGE FLEXION ANGLE RESULTS 21FIGURE 13: GROUP AVERAGE THIGH CORONAL ANGLE RESULTS 24FIGURE 14: GROUP AVERAGE SHANK CORONAL ANGLE RESULTS 25FIGURE 15: GROUP AVERAGE THIGH AXIAL ANGLE RESULTS 27FIGURE 16: THIGH AXIAL ANGLE, INTERACTION AT HEEL STRIKE 28FIGURE 17: GROUP AVERAGE SHANK AXIAL ANGLE RESULTS 29FIGURE 18: GROUP AVERAGE EVENT TIMING RESULTS 31FIGURE 19: FLEXION AND SHANK CORONAL ANGLEPOLAR DIAGRAM 35FIGURE 20: STRAP TENSION AND LEG COMPRESSION FORCES 36FIGURE 21: AXIAL FORCE REACTION TO STRAP TENSION 36FIGURE 22: SUBJECT REACTION 39FIGURE 23: SUBJECT INTERACTION 39viiLIST OF TABLESTABLE 1: GROUP AVERAGE THIGH AXIAL ANGLE MEANS 27TABLE 2: GROUP AVERAGE SHANK A)UAL ANGLE MEANS 291INTRODUCTIONArthritis refers to a group of diseases that involve inflammation of the joints.Technically referred to as rheumatic disease, arthritis can be divided into two types ofillness that differ in their origin: rheumatoid arthritis and osteoarthritis. Rheumatoid arthritisis believed to have a medical origin. A biological or biochemical agent initiates theinflammation and degradation of the afflicted joint. Osteoarthritis (OA), however, has amechanical origin. It is the abnormal loading of forces that initiates the breakdown of thejoint and may cause inflammation. Though “osteo” is Latin for bone, osteoarthritis mainlyaffects the joint cartilage and only during the later stages of the disease are the articulatingbones involved. Inflammation of the joint is also a symptom of osteoarthritis that occursduring the later stages. In the case of the knee joint, one symptom which can be observedduring the early stages of the disease is an abnormal change in the alignment of the lowerlimb.There are two compartments of the femoral tibial joint, one on each side of the knee.Osteoarthritis commonly afflicts only one compartment (especially during the early stages).As the cartilage is mechanically worn away, the afflicted compartment collapses whichcauses the bones to ‘lean” into the narrowed joint space. This process creates the malalignment of the lower limb which in turn increases the already abnormal loads on the joint.Osteoarthritis is characteristic of positive feedback, where the products of the illnessaggravate the illness even further. The patient is trapped in this circular process with thebody unable to cope and in need of intervention.There are many osteoarthritis treatments that deal mainly with the symptoms of thedisease but to be a successful form of intervention, a treatment must deal with the cause ofthe illness: the abnormal force loading. During the later stages of the illness when theabnormal forces are large enough to remodel the articulating bones, surgery can be used toeither remove a wedge from the deformed bones or to completely replace the collapsedjoint with a prosthesis. It is beneficial for the patient to avoid or delay such drastic forms of2intervention whenever possible. One alternative treatment is an osteoarthritis knee bracewhich can be used during the earlier stages of osteoarthritis and is designed to delay orpossibly prevent the progression of the disease.Bracing for osteoarthritis can be considered an intervening form of treatment because itattempts to reduce the abnormal loading across the joint. The theory behind the bracemechanism is that the brace produces an opposing moment or twisting force on the segmentsof the afflicted limb. When the osteoarthritis occurs in the lateral or outside compartment thelimb bends into valgus (see glossary for definitions of valgus and varus). An opposing or varusmoment is required to correct the deformity. This is theoretically possible with a varusinducing knee brace. Similarly, when osteoarthritis occurs just in the medial or insidecompartment of the knee the patients limb goes into varus and requires a valgus knee bracefor correction. Osteoarthritis is much more common in the medial compartment and thusvalgus knee braces are more commonly used. At this time only the valgus knee brace will beconsidered and results from the study of the valgus knee brace are assumed to be applicable tothe varus knee brace.The valgus knee brace is a custom fit orthosis manufactured by a local orthotics firm. It isa modification of a derotation brace used for ligamentous instabilities. It was found thatosteoarthritic patients who wore the brace for a ligament insufficiency also received pain relieffrom their arthritis. The single brace hinge was bent into valgus and moved to the inside of theknee to provide an effective valgus moment. A clinical study (Horlick & Loomer, 1993) wasdone to test the effectiveness of the valgus brace. Pain and function where recorded by bracedsubjects on a daily diary and standing x-rays where conducted to compare the static change inthe femoral tibial angle between brace and no-brace conditions. This double crossover studyfound that the brace provided significant pain relief but there was no significant change inpatient activity level and femoral tibial angle.3Though the study showed clinical significance of pain relief a result very important to anyarthritis sufferer, it did not demonstrate how this was achieved. There is a possibility that thebrace result can be partly explained by the “study effect”, i.e., the tendency of subjects to givea positve pain relief result merely by being a “subject”.The femoral tibia! angle was measured to provide biomechanical evidence of the bracechanging the limb’s alignment. It was not surprising that there was no significant result whenmeasuring the static angle when the subject was standing because: 1) the brace is dynamic innature, a dynamic force strap (see Figure 2 in the glossary) provides the opposing force to thebent hinge arms and is most effective during motion, and 2) forces and moments within theknee are an order of magnitude greater during walking compared to standing and thus itmakes sense to measure the brace effect while the subject is walking. Unfortunately there is noavailable method that can x-ray the subject while they are walking, so the femoral tibial anglewill have to be represented by the outside angle of the leg segments during any dynamic test.With the advent of computerized gait analysis it is feasible to measure the movement of asubject’s limbs while they are walking. The object of this study was to measure the alignmentof the leg while the subjects were not wearing and wearing a knee brace in order to produceobjective biomechanical evidence of the brace effect. This was done by placing reflectivemarkers in locations that represent the axis of the segments. Video cameras tracked themovement of the markers and a computer stored the kinematic data, i.e., the positions of themakers with respect to a reference frame. From the maker positions the segment angle withrespect to a given plane was calculated. The segment angles were illustrated as smoothedcurves averaged from several step cycles.Two choices of reference frames were possible, one fixed with respect to the ground, aglobal reference frame, and one fixed with respect to the subject, a relative reference frame.The frontal angle of the leg alignment was measured with respect to the frontal or coronalplane of the global reference system. The side angle or knee flexion angle was measured withrespect to the sagittal plane of the global reference system. Choosing the frame of reference of4the horizontal angle proved to be more difficult. This was because there was no mention in theliterature on how osteoarthritis affected horizontal rotation and no theory exists on how thebrace might interact in this plane. The segment long axis planes of the relative reference framewas finally chosen because several references related long axis rotation to valgus alignmentchanges. The marker system was redeveloped to measure the long axis rotation of both thethigh and shank (lower leg) segments.Clearly, it was important to discover if the brace actually changes the alignment of theleg. Positive results have a three-fold effect. First, it raises the awareness of a non-surgicalalternative for the treatment of osteoarthritis. Second, it contributes to the understanding ofthe biomechanics, i.e., the forces involved in degenerative diseases. Third, and mostimportant, it can lead to the development and design of a more effective brace, built fromscientific principles rather than using the common trial-and-error methods. Negative resultsmay indicate that the brace affects the loading in the joint in a manner not yet understood.Statement of the ProblemThe valgus knee brace affects the alignment between the thigh segment and the shanksegment in all three planes of motion. Mechanical stress acting in the knee is dependent onthe point of action which in turn is related to alignment of the femur and the tibia. Reducingthe projected angle on the frontal plane between the thigh or shank long axis and thehorizontal, i.e., the coronal angle, would reduce the mechanical stress of the afflicted kneecompartment. In the horizontal plane, internally rotating the shank along its long axis withrespect to the thigh may be beneficial. Changing the alignment of the leg, however, shouldnot be accompanied with a restriction of gait motion in the sagittal plane i.e., the angle offlexion and extension. Showing that the brace has a positive effect on the coronal angle andaxial rotation without restricting gait would provide indirect evidence that the brace is aneffective treatment for OA.5HypothesesThe hypotheses are:1) The valgus knee brace significantly decreases the coronal angles (towards valgus) ofthe thigh and shank segments at four events during the stance phase of gait at eithera medium or fast walking speed.2) The valgus knee brace significantly internally rotates the shank segment along itslong axis and also significantly externally rotates the thigh segment along its longaxis at four events during the stance phase at either walking speed.3) The brace effect of hypothesis 1 and 2 is greater at the fast speed compared to themedium speed, i.e., there is an interaction effect.4) The knee brace does not significantly change the angle of flexion at four eventsduring the stance phase at either walking speed.5) The knee brace does not significantly change the timing of three events during thestance phase (the timing of the first event, heel strike, is defined as zero for allconditions) at either walking speed.Hypothesis 1 stems from the function of the brace to induce a valgus moment about theknee and thus significantly change the alignment of the leg toward valgus. Hypothesis 2 isderived from the theory that a valgus motion should be coupled with an internal rotation ofthe lower leg. Note, in order to twist the shank inward the brace must create an opposingmoment on the thigh and tend to twist it externally. Hypothesis 3 is based on the fact thatbody forces are greater at higher speeds and thus the brace should have an effect thatincreases with gait speed. Finally, hypotheses 4 and 5 state that the brace should not restrictthe range or timing ofmovement in the sagittal plane.6DelimitationsThe delimitations of the study are:1) That patients with OA only in the medial compartment of the left leg were considered;2) That the number of subjects were restricted to 12;3) That only patients with a history of pain while walking were considered;4) That only patients who have reported pain relief and/or improved lifestyle while wearingthe brace were included in the study. This study is to show that the brace alters thealignment of the thigh and shank of patients treated successfully with the brace.7Definition of TermsOsteoarthritis (OA): A degenerative joint disease involving the breakdown of articularcartilage. Believed to be caused by abnormal loading, excessive compressive stress, ormechanical wearing of the cartilage with age. Commonly found in only one compartmentof the knee, symptoms of OA include flattening of the articular surface of the femur,osteophytes, subchondral sclerosis and joint space narrowing of the afflicted compartment.Medial Gonarthosis: OA in the medial compartment of the knee.Genu Varum or Varus: (bow-legged), An orthopaedic term which stems from a Latinword meaning “to bend toward”. In a varus leg, the tibial axis in the frontal plane bends, inrelation to femoral axis, toward the center axis of the body.Genu Valgum or Valgus : (knock-kneed), “to bend away”. In a valgus leg, the tibial axisbends away from the center axis of the body.Coronal Angle of Thigh or Shank: The coronal angle of the thigh or shank is a measureof the lateral angle between the femur and the the horizontal or the tibia and the horizontalin the frontal plane (see Figure 1). It is measured from two of three bony landmarks: thegreater trochanter of the femur, the fibular head and the lateral malleolus. With a knee invarus the coronal angles would be usually greater than 90 degrees and in a valgus knee theangle would be less than 90 degrees.Axial Rotation Angle: The angle of internal or external rotation of a segment about itslong axis. When the segment is vertical (as in standing position), this angle liesapproximately in the horizontal plane.Orthosis: (brace), An exoskeletal device used on the outside ofjoints to stabilize motion.8Valgus Brace: An orthosis used specifically to correct a varus alignment of a knee byproviding a valgus thrust. The valgus brace used in this study is the Generation II OrthoticsUnloader valgus brace. It consists of two semi-rigid plastic shells with supporting straps, asingle high-carbon poli-axial hinge located on themedial side of the leg and a spiralling strap runningfrom the lateral side of the leg (see Figure 2).Three Dimensional (3D) Videography: Ameasurement system that uses at least two videocameras, video equipment, a personal computerand video digitizing software. The system can trackthe motion of a number of reflective markers withina specified area. The computer software cancalculate the displacements, velocities, andaccelerations of the makers and angles betweenthem in all three planes ofmotion.Thigh>990 <900>900ShankFIGURE 1: CORONAL ANGLEFIGURE 2: VALGUS KNEEBRACE9LITERATURE REVIEWBecause OA is a common disease, affecting up to 85 % of the elderly population(Meisel & Bullough, 1984), the volume of literature available is extensive. However, mostof the research is clinical in nature, based on X-ray studies (Cooke, Pichora, Siu,Scudamore, & Bryant, 1989) or long term follow-up studies (Fujisawa, Masuhara &Shiomi, 1979). Much has been written on the biomechanical stress distribution within thejoint using cadavers (Hsu, Garg, Walker, Spector, & Ewald, 1989; Kostuik, Schmidt,Harris, & Wooldridge, 1975; Markolf & Bargar, 1981) or animals (Radin & Paul, 1970;Radin, Parker, Pugh, Steinberg, Paul, & Rose, 1973; Wu, Burr, Boyd, & Radin, 1990).Results of these in vitro studies are inconsistent mainly due to the variety of loadingconditions used. The only reliable way to reproduce true loading conditions is to conductan in vivo study. Both Johnson, Leitle, and Waugh (1980) and Kettelkamp and Chao(1972), conducted research with live subjects and confirmed that static analysis alone is notenough and that dynamic gait analysis would be required to conduct a biomechanical studyof OA. Unfortunately, the majority of gait studies on arthritic subjects to date (Blin,Pailhous, Lafforgue, & Serratice, 1990; Isacson & Brostrom, 1988; Marshall, Meyers &Palmer, 1980; Stauffer, Chao & Gyory, 1977; Suzuki & Takahama, 1979; Waters, Perry,Conaty, Lunsford, & O’Meara, 1987) are clinical in nature. No attempt has been made toanalyze the forces involved, the true origin of the disease. The most comprehensivedescription of the forces involved in OA is provided by Maquet (1984).Maquet details the forces in the frontal plane around the knee joint of normal andarthritic subjects during gait. His model is illustrated in Figure 3. In this plane, the bodyforce, which consists of the weight and inertia of the supported body mass, is denoted bythe vector P. To maintain stability in this plane the body force is balanced by a lateral forceL. The lateral force is believed to be generated by the illiotibial band and supportingligaments. The resultant of L and P is the force R which acts on the plateau of the tibia.Maquet argues that since both compartments of the knee of a normal subject arephysiologically similar and are supported with bone of the same density, the line of action10of the resultant must pass through the center of the knee. The mechanical stress (darkenedarea) is the same for both compartments.For a subject with a varusdeformity, the mechanicalconditions are altered (see Figure4). The line of action of the bodyforce is displaced laterally withrespect to the knee and thus hasa greater moment arm. If thelateral force cannot increase inmagnitude, the resulting forcemust be displaced medially overthe medial compartment. Thisdramatically increases the stressin the medial compartment andcan lead to the formation or theprogression of OA.Unfortunately, there is no known literature on the forces in the transverse plane and itseffect on OA. There are, however, several papers that link the axial alignment of thesegments to the progression and treatment of the disease. The axial alignment and rotationof each segment are reflective of horizontal forces acting in the plane perpendicular to thelong axis. The axial alignment of the tibia is called tibial torsion and is defined as the twistof the shaft between the axis of flexion of the knee and the axis of flexion of the ankle.Kapandji (1987) states that a given static angle of tibial torsion between the two joints isequivalent to an active angle of rotation of the segment. For example, the normalphysiological angle of tibial torsion is 25 degrees and is equivalent to 25 degrees of externalrotation. It has been suggested (Turner & Smillie, 1981, Svenningsen, Terjesen, Auflem, &‘VFIGURE 3: NORMALKNEE11Berg, 1990) that abnormal torsion can lead to osteoarthritis of the knee. However, nodescription was given of the possible mechanism that relates tibial torsion to OA.One method ofmeasurement of tibia! torsion is the angle of foot rotation (Cooke, Chir,Price, Fisher, & Hedden, 1990) in the transverse plane. This method assumes that there isminimal contribution of the internal structures of the foot to rotation and that hip rotation isheld constant. During the standing position it has been shown that the foot rotates inwardafter a positive result of a valgus osteotomety: a surgical correction of the coronalalignment. This is supported by the common knowledge of the inverse case, in whichpatients with medial OA and a varus deformity tend to rotate their foot externally whilestanding as compared to the foot of the normal leg (Loomer, 1993). This relation has notyet been documented or shown in a gait study. However, there appears to be a patternbetween the coronal alignment and the axial alignment of the lower leg. A shift towards avarus alignment tends to be coupled with an external rotation and conversely a shifttowards a valgus direction is coupled with an internal rotation. This pattern is alsodisplayed in young children as they develop their limb alignment. At the age of 2 years,children develop a physiological genu valgum, i.e., valgus alignment and a “protective”toeing-in, i.e., internal rotation (Magee, 1992, Tachdjian, 1972). Thus, if this pattern isgeneralizable, then in the case of an OA patient, it is expected that treatment that includes avalgus change in the alignment of the thigh and shank segments would be also beaccompanied by an internal rotation of the lower limb.There are three current methods of treating OA. Drugs, such as inflammatories, reducepain but do not affect mechanical aspect of the disease (Meisel & Bullough, 1984). Surgerycan be used to correct mechanical alignment by cutting wedges out of the femur or tibia(Maquet, 1984) or by total replacement of the knee with a prosthesis. However, surgery isa traumatic and expensive procedure, not favored among patients. A third treatment of OAis called valgus bracing. It involves the subject wearing an exoskeletal device that stabilizesknee movement. Only recently has the method become widespread in its use.12Very little is written aboutusing a brace that reduces kneestress as a treatment for OA.Cawley, France, and Paulos(1991), provides a detailedliterature survey on bracingresearch but never mentionsany OA braces. Both Baker,VanHanswyk, Bogosian,Werner, and Murphy (1989),and Hoffman, Wyatt, Bourne,and Daniels (1984), describebraces that provide stability inthe frontal plane, however,their possible benefits forarthritis patients is notexplored. Only two types ofarthritis knee braces have beendescribed in the literature. Athird type of brace is reportedby Smith, Juvinall, Corell andNyboer (1970), but it is not a true knee brace because it extends down to the ankle andfoot. Butler, Evans, Rose and Patrick (1983); Cousins and Foort (1975); Dewar, Choderaand Ackerley (1978); and Jawad and Goodwill (1985), all describe a arthritis brace designcalled the CARS-UBC brace or the TVS brace. The design consists of two plastic shells, atelescopic tube assembly and a supporting waist band. All the studies concerning this designwere clinical or descriptive in nature. They merely reported on the number of subjects whocontinued to wear the brace. As reported by Butler, the brace was flimsily designed anduncomfortable to wear. Because no other reports have been written on this type of brace, itis believed not to be in use.13The second type of design is called the valgus knee brace. In 1993 Horlick and Loomerpublished the results of a study done on 79 subjects wearing a valgus brace. The studyshowed that the patients reported significant pain reduction of OA while wearing the brace.It was argued that the brace brought pain relief by reducing stress within the knee. Thisstress reduction is a result of a simple three point loading system that corrects thealignment deformity. Similar loading systems for leg deformities have been mentioned byButler et al. (1983) and Cousins and Foort (1975). The development of the force system fora valgus brace is illustrated in figures 5 and 6. When the brace is built, the hinge is bent invalgus as compared to a varus leg (Figure 5). When the brace is applied (only the hinge isshown, Figure 5 and 6), three brace forces, one on the lateral side and two on the medial,bend the leg into proper valgus alignment. The scientific theory of the valgus brace, is thatthe three point loading displaces the reaction force medially toward the center of the kneeand reduces the stress in the medial side (Figure 6). This reduction in stress will possiblyhalt the progression of OA. It would be very difficult to test this theory because it wouldrequire a device that can measure the contact stress in vivo without affecting thebiomechanics of the knee. However, measurement of the leg alignment with and without abrace would test if the brace is creating a valgus change and thus provide indirect supportthat a three point loading system exists and that there is a shift in the reaction force withinin the knee.14METHODOLOGYSubject SelectionOne group of 12 subjects was recruited from local orthopaedic surgeons using arecruitment form as shown in Appendix A. All the subjects were proscribed a Generation IIUnloader brace before they were considered for the study. The inclusion criteria for the studywere: 1) male, 40-65 years. old; 2) medial OA in the left knee only; 3) has owned a medialhinge Gil Unloader brace more than one month and wears it on a regular basis; 4) has painwhile walking without brace; 5) has pain relief with brace; 6) typical muscle build, not tooathletic; and 7) manual dexterity adequate for self application of brace. The exclusion criteriaare: 1) arthritides other than OA; 2) previous fracture of ipsiateral femur or tibia; 3) previoussurgery to the affected knee other than arthroscopy, debridement, or partial menisectomy; 4)fixed flexion deformity greater than 15 degrees; 5) flexion less than 115 degrees; 6) leg lengthdiscrepancy greater than 2 centimeters; 7) skin disease or peripheral vascular diseasepreventing brace application.EquipmentA total of eleven pilot studies were conducted to develop a knee measurement system forthis study. The system was required to measure knee motion while a subject is not wearing(no-brace condition) or wearing (brace condition) a valgus knee brace. Of particular interestwas the coronal angle of the thigh and shank segments in the frontal plane but the system hadto also measure the sagittal and the axial angles. Thus the system had to be a threedimensional video based system with at least two video cameras or a 3-axis goniometer basedsystem.Goniometer SystemBoth a video based system and a goniometer based system were tested in the pilot studiesbut the goniometer system was rejected because of several fhndamental problems with usingthis device. The main problem of a goniometer based system was the method of attachmentto the leg. The device had to be removed between the no-brace and brace conditions and thus15did not maintain the same relative position with respect to the leg segments. Calibration wasawkward because it was impossible to determine if the angular differences between the twoconditions during the standing (calibration) trials was due to movement of the instrument oran effect of the brace. This is in contrast with the video based system where it was foundout that calibration was not a problem because the markers were located on the skin andwere not disturbed during brace application or removal. Similarly, during the bracecondition, the goriiometer had to be attached directly to the brace and thus it measured thebrace motion, not the leg motion. It must be determined if there is any slippage between thebrace and the leg. Another problem of attachment was that the goniometer tended to slipdown the leg especially during the no-brace condition. This slip was notice by the wearerand was also reflected as an average drift in the output data.Finally, there seemed to be an excessive amount of translation in the sliding devices.With the 3-axis measurement system it was impossible to check if this translation wouldaffect the accuracy of measured rotations. Most of this translation is believed to occurbecause of the misalignment of the goniometer axis with respect to the actual anatomicalknee axis of motion. Theoretically, as long as the goniometer axis and corresponding kneeaxis are parallel, the measurement should be correct, however, the accuracy ofmeasurement decreases as the offset distance between goniometer axis and the knee axisincreases. The goniometer design allows close positioning of the knee flexion axis but notof the varus and shank axis,Video and Treadmill SystemThe final system developed as a result of the pilot trials and used in the study is shownin Figure 7. Two gen-locked Panasonic digital video cameras (Model VW-D5 100) werepositioned at an approximate distance of three meters from a treadmill (QuintonInstruments, Model 24-72). One was located directly in front of the treadmill and the other90 degrees to the side. The motion of the subject was recorded at 60 Hertz and the subjectwas illuminated by lights place above each camera. The aperture on the camera lens was setat 1/500.16A volume of approximately one meter cubed was calibrated by the use of calibrationspider and tripod positioned on the center of the treadmill track. A total of 12 spheres werelocated within the volume and were visible from both cameras. Peak Performance___________________________— (registered) software wasused to digitize the positionsof the spheres videotapedfrom both cameras. ThreeCT dimensional coordinates- —— were calculated using thedirect linear transformationCS method (Shapiro, 1978).The net residual meansquare error between theposition of the spheresFIGURE 8: NO-BRACE FIGURE 9: BRACEt t d in the Ub ti nCONDITION CONDITIONs a e cal ra 0FIGURE 7: EQUIPMENT SETUP17spider manual (relative to the zero coordinate sphere) and the position calculated was foundto be 3.2 mm. The spider was removed and the cameras were fixed in position during thefilming of the trialsSubject SetupEach subject was assessed at an orthotics clinic (Generation II Orthotics) before visitingthe university gait laboratory. At the clinic the subject was checked to insure that they fitthe study criteria and that their brace was fitting properly. Once the subject had understoodthe study protocol and had given written consent they were ready to participate in thestudy.At the laboratory,each subject wasprovided skin tightbicycle shorts andrunning shoes. Reflectivemarkers slightly smallerthan ping pong balls wereplaced on the afflictedlimb of each subject. Fivereflective markers wereplaced on the thigh andshank segments. Themarker locations duringthe no-brace and braceconditions are shown inFigures 8 and 9, Figure10 is a three dimensionalstick drawing of thebraced leg and details theSubject’sThighAnteiiorThighMarkerGreaterTrochanterMarkerSubject’sShankDefinedThigh AxisFibular HeadMarkerDefinedShank AxisLateral• —‘MafleolusMarkerLatera1 Side View)AnteriorShankMarkerFIGURE 10: MARKER Ai1D AXES POSITIONS18marker positions. The thigh segment was defined by the greater trochanter (proximal) andfibular head (distal) markers and the shank segment by the fibular head (proximal) andlateral malleolus (distal) markers. A marker was also placed on the midpoint of the anteriorsurface of each segment. The three dimensional coordinates were determined using thesame method as the calibration spheres. From these data the movements and angles of thethigh and shank segments in a global coordinate system were determined. The knee flexionangle was calculated from the angle projection of the greater trochanter, fibular head andlateral malleolus markers, all in the sagittal plane. The angle projection of the same markersin the coronal plane provided the knee varus rotation. The axial rotation of each legsegment was taken as the relative rotation of the anterior midpoint marker with respect tothe axis of the segment defined by its own proximal and distal markers (see Appendix B forformulas). Footswitches were worn on the bottom of the heel and the big toe to indicate thestart (heel strike, event 1) and the end (toe-ofl event 4) of the stance phase of each stepcycle. The foot switches activated two LEDs within the cameras’ view. The points of lightwere automatically digitized and the event times recorded by the Peak Performance(registered) program.Data CollectionSix types of trials were conducted: two standing trials, one videotaped standing on atreadmill without a brace (no-brace condition: Figure 8) and one videotaped while wearinga valgus brace (brace condition: Figure 9); two medium speed walking trials, no-brace andbrace; and two fast speed walking trials, no-brace and brace. Each trial was filmed for about2 minutes and the order of brace condition was randomized within subjects. The mediumspeed was defined as the normal walking speed of each subject and was calculated as anaverage of ten trials in which each subject walked at a regular pace across a floor, timedbetween two markers, ten meters apart. The fast speed was defined as the maximum speedthe subject walked on the treadmill before naturally breaking into a run.19Data AnalysisFor the walking trials, the data analyzed were the normalized average of five sequentialstep cycles where each cycle is the time from one heel-strike to the next heel-strike of thesame foot. For the standing trials, the static angles were determined as the mean anglevalues from one second (60 frames) of videotape. The three dimensional coordinates weresmoothed with a Butterworth filter at 6 Hz.The walking trial data for the flexion angle during the whole step cycle appearedsimilar to Figure 11. The angle data were statistically analyzed at four events during thestance phase: first event, the onset of heel strike (HS); second event, the first inflectionpoint of the flexion curve, i.e., point of maximum flexion during mid-stance (MAX); thirdevent, the second inflection point and thus the minimum flexion angle during mid-stance(MIN) and fourth event, toe-off (TO). Heel strike and toe-off indicate the onset and thediminution of knee joint compressive stress and thus are critical times in relation to theprogression of OA within the joint. The times of inflection are considered importantbecause they represent the change in direction of knee flexion, the dominating angle of kneejoint motion.Twenty three dependentvariables were used for thestatistical analysis. The first threewere the percent-cycle times ofthe two inflection points and toe-off (MAX, M1N, and TO inFigure 11) and the last twentywere the five defined anglesmeasured at each of the fourevents. The five angles are: 1)the relative flexion angle in thesagittal plane between the thigh6oKneeFlexionAngle 30(degrees)0•HS MAX MLN TOStep Cycle (%)FIGURE 11: EVENT LOCATIONS ON ATYPICAL FLEXION ANGLE CURVE20and the shank; 2) the coronal angle of the thigh, measured from the thigh segment to thehorizontal (CT in Figure 8); 3) the coronal angle of the shank (CS); 4) the axial angle of thethigh, measured in the horizontal plane perpendicular to the long axis of the segment (DE inAppendix B) and 5) the axial angle of the shank. Note, for the walking trials, each dependentvariable was an average of the five step cycles. The average gait angle for the walking trialswas calibrated with respect to the standing position, “no-brace” condition, i.e., the standing,no-brace trial readings was taken to be the reference point for the subject.The study was a repeated measures design with two factors and 23 dependent variables.The factors of interest are brace condition: no-brace and brace, and walking speed: mediumand fast. Thus, there are four conditions defined by combining the levels of the two factors.Each of the 12 subjects was tested at every condition. Therefore, the study is a (12: subject X2 :brace X 2: speed) RM design. The statistical analysis consisted of twenty three separate(12x2x2) ANOVAs all at the p<O.05 significance level.The statistical analysis is defined as follows:- three ANOVAs to test the three event times (MAX, IVUN, TO);- eight ANOVAs to test the thigh coronal angle (CT) and the shank coronal angle (CS) atthe four events;- eight ANOVAs to test the thigh axial angle and the shank axial angle at the four events;- four ANOVAs to test the fiexion angle at the four events.21RESULTSFlexion AngleFigure 12 illustrates the group average flexion angle curves for the speed 1 trials andthe ANOVA results for both speeds. The curves for the speed 2 trial are similar to thespeed 1 conditions. The complete set of graphs and statistics of the group averages and ofthe individual subjects are found in Appendix C and D. Each column of the ANOVA resultstable represents the outcome of a 2 x 2 (2 speed conditions x 2 brace conditions) repeatedmeasures ANOVA, one done at each of the four events during the stance phase. Therewere four conditions defined by combining the levels of the two factors: speed and brace.FLEXION ANGLE ANOVA RESULTSSOURCE HS MAX MIN TOSPEED * *BRACE *SXB60AVERAGE KNEE FLEXiON ANGLE: CALIBRATEDSPEED 1, 12 SUBJECTS, NO-BRACE & BRACE50403020100-10HS 10 20 30 40 50 60PERCENT OF STEP CYCLE—NO-BRACE + BRACE70 80 90 HSkey: * = significant, p <.05FIGURE 12: GROUP AVERAGE FLEXION ANGLE RESULTS22The conditions are defined as: S1B1, medium speed and no-brace; S1B2, medium speedand brace; S2B 1, fast speed and no-brace; and S2B2, fast speed and brace. Completeprintout of the ANOVA tests including the values of the source of error variances and pvalues are included in Appendix C. The test result of the three sources of variances: speed,brace and interaction, are listed in each row of the ANOVA table in Figure 12. An asteriskrepresents a significant test result and a blank space represents a non-significant test result.All tests were done at the p < .05 significance level.The ANOVA table indicates that the source of variance attributed to the bracecondition was not significant at the 1st, 2nd and 4th events but was significant at the 3rdevent (F(lll) = Q, p<.OOl), the point of minimum flexion. The flexion angle means andstandard deviations (in degrees) of the four conditions at the third event are:S1B1 S1B2 S2B1 S2B22.41 4.81 2.12 4.80(4.17) (4.63) (4.73) (4.98)Thus the brace, during the stance phase, had no effect across both speeds on the flexionangle except during minimum flexion where, as apparent by the flexion graph, the braceprevented full extension. The hypothesis stated that the brace would have no effect on theflexion angle across both speeds at each event. This cannot be accepted for the third event,however.In Figure 12, the angle positions of events 2 and 3 of the no-brace and brace conditionare indicated by circled numbers. At event 2, there is no significant difference between theflexion angle of the no-brace and brace condition so they are both indicated by the number2. However, at event 3 there is a significant difference between the no-brace and bracecondition positions. The no-brace condition is indicated by number 3 and the bracecondition by 3’ (three-prime). These angle positions will be referred to in the discussionsection where an explanation will be provided for why there is a significant brace effectduring event 3.23Though not specifically stated in the hypothesis, the flexion angle is expected toincrease with speed (recall that the hypothesis was concerned with the interaction betweenspeed and brace) and the table in Figure 12 indicates that there was significant speed effectonly during the 1St and 2nd event. This suggests that the limit of range of motion during the3rd and 4th event is independent of speed. In other words, at this point flail extension isprevented by static forces, i.e., forces that depend on position and not on movement (therelation between static and dynamic forces and brace design is considered in detail in thediscussion section).As the table indicates, there is no significant effect of the speed x brace interactionduring any of the events. Thus the brace effect does not significantly change across the twolevels of speed and vise-versa. This provides further support that the brace effect seen at the3rd event is a static phenomenon.Thigh Coronal AngleThe ANOVA table in Figure 13 indicates that there was no significant brace effect onthe coronal angle of the thigh. This is contrary to the hypothesis that the brace wouldincrease the thigh coronal angle (i.e., towards valgus) during all four events.Throughout thestep cycle the average range of motion of the thigh in the coronal plane is within +1- 3degrees from the calibration position. Given this limited range of motion and the fact thatthe thigh segment has several large abductor and adductor muscle groups that cross the hipjoint, it is possible that any brace effect on the thigh would be too small to detect. Anybrace effect is more likely to be seen on the lighter shank segment.As gait speed increases, the forces and moments on the segments increase as well. It isexpected that there should be an increase in the thigh coronal angle as the varus forcesincrease with speed. As seen in Figure 13 there is a significant speed effect only during the3rd event. This shows that the range of motion of the thigh segment is limited even byspeed. Since the thigh is connected directly to the pelvis, thigh motion in the coronal plane24THIGH CORONAL ANGLE ANOVA RESULTSSOURCE HS MAX MIN TOSPEED *BRACESXBmay be limited to reduce swaying of the center of mass. This would be most important duringthe weight transfer phases of gait just after heel-strike (1St event) and just before toe-off (4thevent). Because there is no interaction effect, the effect of speed during the 3rd event does notchange with brace condition.Shank Coronal AngleAcross both speeds and all events, shank coronal angle means for the brace condition werefound to be less than the corresponding mean of the no-brace condition. However, as shownin Figure 14, only during the fourth event was the brace effect significant (F(111)p=.O24). The hypothesis stated that the brace would significantly reduce the shank coronalangle at all four events. But this hypothesis can be only accepted at the toe-off event.10AVERAGE THIGH CORONAL ANGLE: CALIBRATEDSPEED 1, 12 SUBJECTS, NO-BRACE & BRACEr,)00-5HS 10 20 30 40 50 60 70 80 90 HSPERCENT OF STEP CYCLE_NO-BRACE + BRACEkey: * = significant, p <.05FIGURE 13: GROUP AVERAGE THIGH CORONAL ANGLE RESULTS25SOURCE HS MAX MIN TOSPEED * *BRACE *SXBThe shank coronal angle means and standard deviations (in degrees) of the four conditionsat the fourth event are:The angle position of events 2 and 3 of the no-brace and brace condition in Figure 14 areindicated by circled numbers, similar to Figure 12. These positions will be used in thediscussion to explain the relationship between the flexion angle and the shank coronal angle.10AVERAGE SHANK CORONAL ANGLE: CALIBRATEDSPEED 1, 12 SUBJECTS, NO-BRACE & BRACE5Cl)HS 10 20 30 40 50 60 70 80 90 HSPERCENT OF STEP CYCLE—NO-BRACE + BRACE-5++++++SHANK CORONAL ANGLE ANOVA RESULTSkey: * = significant, p <.05FIGURE 14: GROUP AVERAGE SHANK CORONAL ANGLE RESULTSS1B1 S1B2 S2B1 S2B26.619 5.310 7.387 6.951(3.51) (3.08) (3.91) (4.44)26The first row of the ANOVA table indicates that the speed effect was only significant atthe onset and end of the stance phase (HS and TO). This pattern was opposite to what wasobserved with the thigh coronal angles where there was a speed effect during the middle ofthe stance phase. This indicates that the thigh and shank segments play different roles in speedcontrol during the stance phase of gait. Perhaps the thigh position is consistent at the onsetand end of the stance phase to control the center of mass but the shank angle varies withspeed to allow proper foot positioning during weight transfer. The relationship between theflexion angle and the shank coronal angle and their possible connection to static forces isexplored later in the discussion section.There was no interaction between the speed and the brace effect during any of the events.This is most surprising at the fourth event where there is a significant speed and brace effect.This indicated that at this point the speed effect and the brace effect operated with twoseparate processes. The brace effect is mostly due to static forces and the speed effect isdynamic. In other words, the brace would provide about the same correction in angle at thisposition, if the leg was moving at any speed. The difference between the no-brace and braceangle is constant with speed.Thigh Axial AngleThe results of the statistical tests on the thigh axial angles in Figure 15 are most promising.There was a significant interaction effect between speed and brace during the first two eventsand a brace main effect across all the events.In the case of the first event, heel strike, the significant interaction effect (F(ll 1) =<.001) is illustrated in Figure 16. The effect of the brace is greater at the fast speed (shownas delta SP2) as compared to the effect at the medium speed (shown as delta SP1). This wasthe same pattern seen for the second event (interaction F(11 1) = , =.04 1) and the thirdevent (interaction F(111)= 4, =.055). Also, on average across both speeds, the thigh axialangle for the brace condition was greater (externally rotated) than the no-brace condition forall four events.27SOURCE HS MAX MIN TOSPEED * *BRACE * * * *SXB * *EVENT SIBI SIB2 S2BI S2B21) HS -10.7 0.3 -16.3 -2.72) MAX -3.5 4.5 -5.9 4.03) MIN -0.4 4.2 -0.5 584)TO 5.7 9.7 6.2 10.920AVERAGE THIGH AXIAL ANGLE: CALIBRATEDSPEED 1, 12 SUBJECTS, NO-BRACE & BRACE10C,’0-10-20ITS 10 20 30 40 50 60 70 80 90 HSPERCENT OF STEP CYCLE—NO-BRACE + BRACETHIGH AXIAL ANGLE ANOVA RESULTSkey: * = significant, p ‘C .05FIGURE 15: GROUP AVERAGE THIGH AXIAL ANGLE RESULTSTHIGH AXIAL ANGLE MEANS (degrees)TABLE 1: GROUP AVERAGE THIGH AXIAL ANGLE MEANS28It is clear that the brace creates significant external rotation forces on the thigh (and thusinternal rotation forces on the shank). It can be argued that the large change in the thigh axialangle was simply due to soft tissue movement on the anterior aspect of the thigh. If this wasthe case then there would be a constant “shift” between the curves of the no-brace and bracecondition throughout the step cycle. However, on an individual basis (see subject graphs,Appendix D, pages 75 & 76), most subject graphs showed a convergence of the no-brace andbrace curves either near the end of the stance phase or during the swing phase.The speed effect was only significant for the first half of the stance phase. Thus, at thebeginning of the stance phase the thigh axial angle is affected by the brace condition and thegait speed with static and dynamic forces working together but near the end of stance theangle is affected only by static forces (speed independent) of the brace condition.Shank Axial AngleAs shown in the ANOVA table in Figure 17, the brace effect was significant in internallyrotating the shank axial angle during events 1 (F(llj) =,p=.O12) and 3 (F(111) ==.O45). What was surprising was that the brace had an opposite effect (external rotation) atevents 2 (F(111) = 4A, =.O59) and 4 (F(1,1)= , =.O15) contrary to that stated by the— no-brace++++ braces SP2THIGH 0A)ALANGLE(degrees)-15FIGURE 16: THIGH AXIAL ANGLE, iNTERACTION AT HEEL STRIKEMed. Speed Fast Speed29SHANK AXIAL ANGLE ANOVA RESULTSSOURCE HS MAX MIN TOSPEED * *BRACE * *SXBEVENT SIBI SIB2 52B1 52B21) HS -1.8 -4.5 -2.0 -4.82) MAX -1.4 -0.2 -2.2 -0.13) MIN -0.5 -3.5 0.7 -1.84) TO 7.2 9.8 10.4 13.6252015AVERAGE SHANK AXIAL ANGLE: CALIBRATEDSPEED 1, 12 SUBJECTS, NO-BRACE & BRACE105-5-10HS 10 20 30 40 50 60 70 80 90 555PERCENT OF STEP CYCLE_NO-BRACE + BRACEkey: * = significant, p <.05FIGURE 17: GROUP AVERAGE SHANK AXIAL ANGLE RESULTSSHANK AXIAL ANGLE MEANS (degrees)TABLE 2: GROUP AVERAGE SHANK AXIAL ANGLE MEANS30hypothesis. It was expected that the shank axial angle would be externally rotated across allfour events in the reverse direction of the thigh axial angle and this effect would also beexpected to be greater on the shank because it is lighter than the thigh segment. However,note that the shank had the correct anticipated effect only when the knee was filly extended(events 1 and 3), i.e., when the shank long axis is in line with the thigh long axis. Thus thebrace internally rotated the shank in reaction to the external rotation of the thigh at fUllextension but when the knee was flexed the axial rotations of the segments were not mutuallydependent. The discussion section introduces other mechanisms that can explain the behaviorof the shank axial rotation.Referring to the ANOVA table in Figure 17, there was a speed effect on the shank axialangle only during the last half of the stance phase. This was opposite to the thigh axial angle inwhich there was a speed effect only during the first half of stance. Also indicated by thecoronal angles, the thigh and shank segments play different roles in speed control during thestance phase of gait. The shank axial angle controls foot rotation which is sensitive to push-offforces that increase with speed during the last half of the stance phase. However, during thefirst half the stance phase the speed dependent braking forces act though the base of the heeland are essentially independent of foot rotation.As with the shank coronal angle, there was no interaction between the speed and the braceeffect on the shank axial angle during any of the four events events. This provides moreevidence that the speed effect and the brace effect operate with two separate processes, onestatic and the other dynamic.31SOURCE MAX MIN TOSPEED * * *BRACESXBkey: * = significant, p c .025FIGURE 18: GROUP AVERAGE EVENT TIMING RESULTSEvent TimesThe ANOVA table in Figure 18 shows the test results for the timing of events 2, 3 and 4(see Appendix C, pages 72 & 73 for values). Event 1 (HS) always occurs at time zero. Thereis no significant brace effect on any of the three event times. Thus the brace does not hasten ordelay the percentage times of maximum flexion and minimum flexion during the stance phaseand there is no difference of the percentage length of the stance phase between the no-braceand brace condition. This result was predicted by the alternative hypothesis.The event times did, however, change with speed. In terms of percentage of step cycle,event 2 timing increased with speed thus increased braking time, and the event 3 and 4 timingdecreased with speed which hastened the push-off phase and reduced the time the foot is onthe ground (when in contact, the foot is not moving with respect to the ground and canrestrict the speed of the total body motion). There was no significant interaction between thespeed and brace effects and thus the brace did not interfere with the adaptive changes of eventtiming with speed.EVENT TIMING ANOVA RESULTS32Results SummaryThe results of the study can be summarized by six statements.There was an expected significant interaction effect between brace and speed on:1) the thigh axial angles during the first two events(heel strike and max. flexion).The brace had an expected significant effect on:2) the shank coronal angle during event 4 (toe-off);3) the thigh axial angle during all the events;4) and on the shank axial angle during events 1 (heel-strike)and 3 (mm. flexion).The brace had an unexpected significant effect on:5) the flexion angle during event 3 (mm. flexion);6) and on external rotation of the shank axial angle during event 4 (toe-off).33DISCUSSIONThe brace has an effect on the segment angles but not in the same way as hypothesized.There was no overall shift or increase in the segment angles in the predicted direction. Whatthis study has shown is that the complete effect of the brace can not be explained by a simplesubject/brace reaction. There is an interaction between the brace and the subject which iscomplex in nature.Across a group of subjects and across two walking speeds, this study demonstrated thatthe brace did have a significant effect on the segment angles of the thigh and shank. Themechanical process introduced in the literature stated that the subject simply reacts to a threepoint bending force. However, the results obtained were not consistent with a three pointbending force reaction (referred to as the subject/brace reaction). It is possible that additionalbrace forces and the process of subject/brace interaction can explain the results of this study.The idea of the subject/brace reaction does not have to be abandoned all together but canbe redeveloped and incorporated into a more complex process. Thus, the subject/bracereaction does explain some elements of the observed behavior of the brace effect. Briefly, thesubject/brace reaction accounts for the static forces that the brace provides and the valgusbending forces generated (or transmitted) by the brace essentially depend only on the legposition. If the valgus force was dynamic it would depend directly on motion. It should benoted that static and dynamic are actually abstract terms and sometimes there is a grey regionthat separates their respective definitions. For example, the bolt shear force that holds thebrace hinge to the shells is a static force because there is no relative motion between the parts,but it can also be considered dynamic because it allows them to accelerate together.In the case of the brace, the valgus bending force is primarily produced by brace iron (orhinge arm) stress and strap tension. The brace iron force (IR) is related to the relativedifference between the brace alignment and the the leg alignment and thus it is static and itdepends on position rather than motion. The tension in the strap (ST) also depends on the34relative position of the thigh and shank brace shells. The strap tension is maximum when theleg is fully extended. Thus, in all, the valgus bending force is primarily a static force. There arealso dynamic forces in the brace design such as strap elastic strain (SE) and hinge axisalignment force (HG). Both static and dynamic brace forces can be used to explain the studyobservations. Static forces are useful in quasi-static analysis, where the system is assumed tobe stable at any point in time (freeze-frame), and dynamic forces are useful when the directionofmotion or type ofmotion is important.Flexion Angle and Shank Coronal AngleIn the results section the segment angles in each plane were discussed separately.However, there are relationships between angles in separate planes. For example, there is arelation between the flexion angle and the shank-coronal angle. This connection is not onlydue to knee geometry, supporting structures and muscle forces (which are assumed to befactored out when comparing no-brace and brace conditions), but also due to the brace designitself Relating the coronal angle to the corresponding flexion angle may explain why the bracesignificantly prevented full extension during event 3 but had no significant effect on the shankcoronal angle.The portion of the step cycle between events 2 and 3 is shown in Figure 19 as a polardiagram, where the shank coronal angle is plotted as a function of the flexion angle. The anglepositions 2, 3, and 3’ are the same positions indicated in flexion angle graph in Figure 12 andthe shank coronal graph in Figure 14. The polar graph contains the same information betweenevents 2 and 3 as the graphs in Figure 12 and 14 except the time component is removed. Thedirection of motion is indicated by arrows on the curves. During the no-brace condition (solidline) the knee extends from position 2 to the point of minimum flexion at position 3 while theshank-coronal angle goes into valgus. The brace condition (dashed line) starts from the samepoint at event 2 but fails to reach full extension at event 3 (indicated as position 3’). Note thatthe brace condition maintained the same valgus change as the no-brace condition. The arrowsindicated by the the number 4 represent the brace forces required to divert the path of motionfrom the no-brace condition between events 2 and 3. If the brace condition path of motion35was extrapolated past point 3’ to fi.ill extension (dotted line), it would reach point 3”, a pointvalgus to the no-brace condition. The larger arrows shown by number 5 represent theadditional brace forces required to produce the hypothetical motion.Figure 19 shows that it is possible for the brace to produce valgus bending forces withouta significant change in coronal angle. Note that the brace forces, 4 and 5, resist extension aswell as create a valgus movement and this is consistent to the properties of the strap tensionforce. The diagonal strap wraps around the leg in a helical fashion (see Figure 20). As theknee extends, the ends on the helix are moved further apart and the strap tension (ST) resistsextension. At the same time the strap applies compression (C) on the lateral aspect of theknee. This is the same maimer in which a lengthened cylinder of a constant volume will haveits diameter compressed. The strap mechanism actually acts like a machine that transfersmuscle force in the sagittal plane to compression force in the coronal plane. The lateralcompression force becomes the essential component in the three-point loading system inFigure 20 that produces a valgus bending force.IVARUSShankCoronalAngleVALGUSip,0EXTENSION FLEXIONFlexion Angle+FIGURE 19: FLEXION AN]) SHANK CORONAL ANGLE POLAR DIAGRAM36Knee geometry and the supporting structures resist any valgus deviation for a givenflexion angle. Thus, it may be possible to apply large valgus bending forces without asignificant valgus angle change. Additional valgus forces (number 5 in Figure 19) that cancreate an significant change in valgus angle may be not necessary or may be uncomfortable forthe subject. The results of the shank coronal angle analysis shows a valgus trend across all theevents and both speeds, but only during the fourth event (toe-off) was was there a significantchange due to the brace condition. The knee may be flexed enough during toe-off to relax thesupporting structures and allow valgus laxity.Thigh Axial Angle and Shank Axial AngleThere is another consequence of the helical structure of the strap. A force component ofthe strap tension (vector A in Figure 21) produces a twisting moment about the long axis ofeach segment. When the leg is fully extended as in Figure 21, the strap tension reactioncreates an external moment about the thigh long axis and an internal moment about the shankSTA(ext)ST£/2STØL)FIGURE 21: AXIAL FORCE REACTIONTO STRAP TENSIONFIGURE 20: STRAP TENSION AND LEGCOMPRESSION FORCES37long axis. These moments are equal in magnitude and opposite in direction. Because thesemoments are dependent on strap tension they are maximum at full extension. The strapmechanism explains the significant brace effect on external rotation of the thigh segmentacross all the events and the significant internal rotation of the shank segment during events 1and 3: points of full extension. However, it does not explain the external rotation of the shankduring events 2 and 4.The rotation moment analysis assumes that the thigh and shank share the same axis ofrotation. However, as soon as the knee is flexed, the thigh and shank axis are not in line andthe axial moments do not have to balance. During flexion, a separate mechanism may beacting on the shank axial rotation and have no significant effect on the thigh axial rotation.This mechanism may be the dynamic force of hinge rotation. The brace hinge is designed toallow controlled valgus/varus rotation along with flexion/extension. As soon as the kneeflexes, a component of valgus/varus rotation acts in the direction of the long axis of the shankand thus creates internal/external rotation. The magnitude and the direction of the hinge axialrotation force depends on the type of motion and thus is defined as a dynamic force. Thehinge dynamic force creates external rotation during flexion and internal rotation duringextension. Thus, during events 1 and 3 the strap and hinge forces work together and createinternal rotation on the shank and during events 2 and 4 the hinge creates an external forcewhich overcomes the strap internal force which is reduced during flexion.The results showed a greater change in angle due to the brace main effect on the thighaxial angle compared to the shank axial angle. This is contrary to the assumption that theshank would have a lower resistance to rotation because it is a lighter segment. However,during the stance phase the shank and foot complex are essentially fixed to the ground andhave large shearing forces at the point of contact. Also, the ankle and forefoot are resistant torotational forces along the long axis of the shank. In contrast, with the thigh segment, thesupported body is free to rotate about the long axis during the single support phase providingonly inertial resistance to thigh axial rotation. Also, during the double support phase when thebody is fixed, the hip joint allows freedom ofmotion about the thigh long axis. Thus, when the38boundary conditions of the segments are considered, the heavier thigh segment has a lowerresistance to axial motion during the stance phase compared to the shank segment.Subject InteractionOne of the main assumptions of the subject/brace reaction is that muscle forces do notchange between the no-brace and brace condition. In other words, there is no subject/braceinteraction: the subject does not change their muscle firing pattern in response to wearing abrace. A process incorporating the feedback and anticipatory effects of a subject/braceinteraction would be complex in nature compared to the subject/brace reaction.Figure 22 is a systematic flow diagram of the subject reaction process: a combination ofthe subject/OA reaction and the subject/brace reaction. The diagram is concerned only withthe effects on the coronal and axial angles. A mirror image stick figure is located under each“state” of the process and represents the alignment of the lower limb. Each state in the subjectreaction process is assumed to be stable and does not change unless there is ftirtherprogression of the disease or there is some form of intervention. During the subject/OAreaction, OA causes both a varus change in the coronal plane and external rotation along thelong axis creating the familiar bow-legged appearance. The circled connection between thecoronal and axial arrows in Figure 22 indicates that the two processes are coupled, one doesnot occur without the other. The subject/brace reaction is also coupled and simply corrects thevarus alignment and the rotation back to normal: it reverses the OA effect.The study results do show support for the subject/brace reaction process. The brace haslittle effect on the sagittal plane motion while a large effect on axial motion. If the subjectsinteracted with the brace and changed their muscle firing pattern there would be an expectedchange of movement pattern in the sagittal plane. Most of the leg’s muscles are dominate inthe sagittal plane where the gait forces are an order of magnitude greater than compared tothe other planes. The brace, however, is designed to transmit forces to the axial and coronalplanes with little resistance in the sagittal plane. So, it is possible that the subjects are justreacting to brace forces. If an interaction process does exist, it would have to be complex inorder to affect axial and coronal motion without disturbing the sagittal pattern ofmotion.FIGURE22:SUBJECTREACTIONCoronal:NormalAxial:MirrorImageStickFigureNormal )(OABraceFIGURE23:SUBJECTINTERACTIONOACoronal:Normal—*VarusAxial:Normal )(SubjectBrace>LessVarus—NormalJ.ALH1d1Subject>LessVarus‘1’Norma1/()(40However, the subject interaction process should be considered because the simple subjectreaction process has many unanswered questions. The foremost question is: how does OAcreate an external rotation? There is a mechanism introduced in the literature section thatexplains how OA affects the coronal angle, but there are no theories as to how OA directlyaffects rotation. It it also not clear how the brace can reverse the OA effect in both the varusand the axial planes at the same time. In addition, the brace would not be able to correct anyrotation about the hip that may be caused by OA because the brace does not cross the hipjoint. The process also assumes that the direction toward normality is always correct, i.e., thatthe best leg alignment for an OA subject would be the same as a subject with no OA. Finally,it is reasonable that a subject will change their walking pattern in response to the disease or atreatment of the disease.The subject/OA and subject/brace interaction is included into the subject interactionprocess shown by Figure 23. In this process, OA only causes the varus deformity and uses themechanism previously explained. This state, however, is not stable and is changed by thesubject/OA interaction. The interaction is a response of the musculoskeletal system (andhigher levels) to the abnormal loading condition caused by the disease or secondary effects.The “motivation” of the interaction is to 1) improve stability, 2) reduce pain and 3) retard thepremature degradation process. There are consequential side effects and thus limitations ofthe interaction process. The limitations are related to the location and strength of musclegroups, geometry of the joints, and the fact that the three motivating forces may be competingfor the same resources and acting against each other. The optimum interaction takes time forthe musculoskeletal system to develop. The subject interaction cannot be a simple “correction”of the deformity. If that was the case, the deformity would never occur.The subject/OA interaction consists of an external rotation of the shank/foot and areduced varus deformity (valgus shift). By externally rotating the foot, the subject can createan offset distance between the ground force vector and the body force vector in the coronalplane. The resulting moment is a valgus bending force about the knee which can decrease thevarus deformity. This benefit can only occur during the push-off stages of the stance phase41and is limited by side effects such as unwanted bending forces in other planes. Note that theexternal rotation is in the direction away from normality, a violation of the normalityassumption in the simple theory.The result is a stable state with reduced varus deformity but at the cost of a normal footalignment. Using the process explained in the literature section, page 10, a valgus brace canreduce the varus deformity even more, perhaps to the normal leg alignment. In the complextheory shown in Figure 23, the brace is considered to have no consequential effect on rotation.This subject interaction process is the other extreme of the reaction process, in which OA andthe brace have the fIjll effect on rotation rather than no effect.As with the OA effect, the subject interacts with the brace effect. Essentiality, thesubject/brace interaction is the reverse process of the subject/OA interaction. The subjectinternally rotates the limb and consequently the leg goes into varus. In this case, the coronalangle increases in the direction away from normality. The advantage of the subject/braceinteraction is that the subject regains the original axial alignment. Thus the combined effect ofthe brace effect and the subject/brace interaction is a change in the axial angle. The bracemakes it possible to retain the benefit of varus reduction from the subject/OA interactionwithout the need for external rotation of the foot, The subject interaction process explainshow the mechanisms introduced in the literature section, (OA increases varus, page 10, bracedeceases varus, page 12) can produce the results found in the study (no brace effect on varusangle, page 24 & 25, large brace effect on axial rotation angle, page 27 & 29).The discussion sections previous to the subject interaction process have explained how thevalgus brace can (without subject/brace interaction) directly affect axial rotation and whythere was little valgus change in the coronal angle. This suggests that the brace valgus effectshown in Figure 23 can also be coupled with an axial rotation. It is also possible, yet notunderstood, that OA can directly effect the axial rotation angle. Several references in theliterature section link axial rotation to the progression of the disease. In all, the true behaviorof the whole subject/OA/brace system probably lies somewhere between the subject reaction42FUTURE RESEARCHThe greatest problem with the subject interaction process is that it is not in a form of atestable theory. This is a problem that is common with many complex processes that describesystems that involve feedback and anticipation. For example, in this case, it is impossible tomeasure the subject interaction when it is hidden within the disease or treatment effect. Thesubject interaction process must be broken down into theories with known mechanisms andmeasurable parameters. The largest task is to determine the correct mechanisms and therelevant parameters. For example, the idea of flight is a complex theory, but it was the Wrightbrothers who recognized the importance of air pressure and the mechanism of controllingwing warping. There are many possible parameters with OA, such as knee laxity, initial varusangle, tissue compliance, etc., but they are not usefhl without a theory that determines theirrole.Complex theories do not describe systems that are random and impossible to understand.Upon examination, patterns will emerge and the system will become predicable within somecertainty. The next step in this research is to study the individual gait data of the subjects andfind patterns across the 12 subjects or across a smaller group that can be related to a relevantparameter. For example, the brace effect may be greater for subjects who have worn the bracefor a long time or who participate regularly in sports. This type of analysis was beyond thescope of this thesis but can be used to determine subject selection criteria for fhture studiesand lay the foundation of the creation of new theories.What is lacking and is needed for further research are new theories on OA and valgusbrace treatment. This study has found that the simple processes introduced in literature sectionare not adequate for describing the system behavior. Research is required to study the relation,if any between OA and axial rotation. The rotation shearing forces within the knee joint mayplay an important role in the progression of the disease. A new theory also has to explain theeffect OA (and the brace) has in the horizontal plane, a plane that is part of the globalreference system and is perpendicular to the sagittal and coronal planes. This plane was not43used in the present study because there was no information or concept on how OA affectedhorizontal movement other than along the long axis of the thigh and shank segments. Theseaxes lie in the horizontal plane only when the knee is extended and the leg is vertical. Thecurrent marker system used in this study, though useful in measuring axial rotation, wouldhave to be modified to measure horizontal rotation.The subject/OA and the subject/brace interaction needs to be defined more clearly. Amethod must be developed to determine if an effect on motion is due to subject response or adisease or treatment effect. It may be impossible to separate the two processes and it can beargued that they are one and the same. However, fundamentally the subject response is acharacteristic of the individual and the disease or treatment effect is common to all affectedindividuals (though not necessary to the same degree). This shows the importance in findingthe relevant individual parameters that can describe the subject interaction.Introducing the concept of subject interaction also brings into question the effect thatdisease and treatment methods have on the other joints of the lower limb. Though medial OAand the valgus brace only directly effect the knee joint, many muscles that cross the knee arebi-articular and thus control movement of other joints such as the hip and anide. Subjectinteraction can change the loading conditions of other joints, sometimes in a negative way, inresponse to the situation at the knee joint. For example, some patients with medial OA in oneknee complain of frequent back pain. It is believed that the pain is due to the abnormal posturethe musculoskeletal system creates by trying to redistribute stress away from the afflictedlimb. These ideas can only be tested by developing theories that explain how the body adaptsto the illness. Future studies can then measure the effect OA and the valgus brace have onother joints of interest.The valgus brace design used in this study is more complicated than a simple legbending device. There are questions as to what effect the dynamic properties of the brace suchas hinge axis rotation and strap elasticity have on the treatment of OA. Further analysis of thebrace can lead to a better prediction of its behavior and new ways of improving its design.44Though forces and moments were not measured in this study, force analysis was used toexplain all the observed motion. It is obvious that once the kinematics have been understoodthe next step would be to measure the forces and moments directly. The same theories thatdescribe the observed motion, if correct, should also be used for any kinetic analysis.45CONCLUSIONThe purpose of the study was to test if the valgus brace changed the alignment of the legaccording to a valgus bending force theory. The results showed that the brace did have asignificant effect but not in the way anticipated. The brace had no effect on the thigh coronalangle but did show a valgus trend on the shank coronal angle which was only significant attoe-off. The brace also prevented full extension during mid-stance but this was believed to becaused by resisting valgus forces in the coronal plane transmitted through the brace’s helicalstrap. This same strap is also believed to be responsible for axial forces that created thesignificant external rotation of the thigh axial angle throughout the stance phase and theexternal rotation of the shank axial angle during knee flexion. Dynamic forces of the bracehinge may explain the unexpected internal rotation of the shank axial angle that accompaniedknee extension. The possibility that the subject interacted with the brace to produce some ofthe observed motion was introduced. The process requires further development before it is atestable theory but it does stress the importance of discovering relevant parameters thatdescribe the individual’s response to the disease or treatment. Further research should includetesting the effect that OA has on horizontal movement and measurement of brace forcesduring gait.46REFERENCESBaker BE, VanHanswyk E, Bogosian SP, Werner FW, & Murphy D (1989). The effect of kneebracing in lateral impact loading of the knee. The American Journal of Sports Medicine, 17,182-186.Bun 0, Pailhous J, Lafforgue P, & Serratice G (1990). Quantitative analysis of walking inpatients with knee osteoarthritis: a method of assessing the effectiveness of non-steroidal anti-inflammatory treatment. Annals of the Rheumatic Diseases, 4, 990-993.Butler PB, Evans GA, Rose OK, & Patrick JR (1983). A review of selected knee orthoses.British Journal ofRheumatology, , 109-120.Cawley PW, France EP, & Paulos LE (1991). The current state of functional knee bracingresearch. The American Journal of Sports Medicine, 1, 226-233.Cooke TDV, Chir B, Price N, Fisher B, & Hedden D (1990). The inwardly pointing knee: anunrecognized problem of external rotation malalignment. Clinical Orthopaedics and RelatedResearch, Q, 56-60.Cooke TDV, Pichora D, Siu D, Scudamore RA, & Bryant JT (1989). Surgical implications ofvarus deformity of the knee with obliquity of joint surfaces. Journal of Bone & Joint Surgery,71B, 560-565.Cousins S & Foort J (1975). An orthosis for medial or lateral stabilization of arthritic knees.Orthotics and Prosthetics, 2, 2 1-26.Dewar M, Chodera JD, & Ackerley K (1978). Clinical trial and development of the CARSUBC knee brace (Report 1978: pp.88-95). Roehampton, London: Biomechanical Research andDevelopment Unit.47Fujisawa Y, Masuhara K, & Shiomi S (1979). The efiuect of high tibial osteotomy onosteoarthritis of the knee. Orthopedic Clinics ofNorth America, !Q, 585-608.Hoffman AA, Wyatt RW, Bourne MH, & Daniels AU (1984). Knee stability in orthotic kneebraces. The American Journal of Sports Medicine, , 371-374.Horlick S & Loomer RC (1993). Valgus knee bracing for medial gonarthrosis. Clinical Journalof Sport Medicine, , 251-255.Hsu HP, Garg A, Walker PS, Spector M, & Ewald FC (1989). Effect of knee componentalignment on tibial load distribution with clinical correlation. Clinical Orthopaedics & RelatedResearch, 135-44.Isacson J & Brostrom LA (1988). Gait in rheumatoid arthritis: an electrogoniometricinvestigation. Journal ofBiomechanics, 451-457.Jawad A & Goodwill C (1985). TVS brace in patients with rheumatoid arthritis orosteoarthritis of the knee. British Journal ofRheumatology, , 416-417.Johnson F, Leitle S, & Waugh W (1980). The distribution of load across the knee, acomparison of static and dynamic measurements. The Journal of Bone and Joint Surgery, 62B,346-349.Kapandji IA (1987). The Physiology of the Joints, Volume 2, Lower Limb. New York:Churchill Livingstone.Kettelkamp DB & Chao EY (1972). A method for quantitative analysis of lateral compressionforces at the knee during standing. Clinical Orthopaedics & Related Research, , 202-213.48Kostuik JP, Schmidt 0, Harris WR, & Wooldridge C (1975). A study of weight transmissionthrought the knee joint with applied varus and valgus loads. Clinical Orthopaedics & RelatedResearch, !Q, 95-98.Loomer RC (1993). By personal conversation.Magee DJ (1992). Orthopedic Physical Assessment. Philadelphia: W.B. Sauders Company.Maquet PG (1984). Biomechanics of the knee (2nd Ed.). New York: Springer-Verlag.MarkolfKL & Bargar WL (1981). The role ofjoint load in knee stability. The Journal ofBoneand Joint Surgery, A. 570-585.Marshall RN, Meyers DB, & Palmer DG (1980). Disturbance of gait due to rheumatic disease.Journal ofRheumatology, 2, 617-623.Meisel Al) & Bullough PG (1984). Atlas of Osteoarthritis. Philadelphia: Lea & Febiger.Radin E & Paul 1(1970). Does cartilage compliance reduce skeletal impact loads? Arthritis andRheumatism, j, 139-144.Radin E, Parker HG, Pugh 1W, Steinberg RS, Paul IL, & Rose P.M (1973). Response ofjointsto impact loading - III: Relationship between trabecular microfractures and cartilagedegeneration. The Journal ofBiomechanics, , 5 1-57.Shapiro R (1978). Direct linear transformation method for three-dimensional cinematography.The Research Quarterly, 4, 197-205.Smith EM, Juvinall RC, Corell EB, & Nyboer VJ (1970). Bracing the unstable arthritic knee.Archives ofPhysical Medicine and Rehabilitation, , 22-28.49Stauffer RN, Chao EY, & Gyory AN (1977). Biomechanical gait analysis of the diseased kneejoint. Clinical Orthopaedics & Related Research, 246-255.Suzuki K & Takahama M (1979). Gait patterns of the diseased knee joint. Journal of JapaneseOrthopedic Association., , 847-853.Svenningsen S, Terjesen T, Auflem M, & Berg V (1990). Hip rotation and in-toeing gait.Clinical Orthopaedics and Related Research, i, 177-182.Tachdjiuan MO (1972). Pediatric Orthopedics. Philadelphia: W.B. Sauders Company.Turner MS & Smillie IS (1981). The effect of tibial torsion on the pathology of the knee. ThcJournal ofBone and Joint Surgery, 396-398.Waters RL, Perry J, Conaty P, Lunsford B, & O’Meara P (1987). The energy cost of walkingwith arthritis of the hip and knee. Clinical Orthopaedics & Related Research, j4, 278-284.Wu DD, Burr DB, Boyd RD, & Radin EL (1990). Bone and cartilage changes followingexperimental varus or valgus tibial angulation. Journal of Orthopaedic Research, , 572-585.50APPENDIX ASUBJECT RECRUITMENT FORMANDSUBJECT CONSENT FORM51SUBJECT RECRUITMENTProject Title:The effects of Valgus Bracing on the 3D Kinematics of Gaitin Patients with Osteoarthritis of the knee.Dear {Orthopaedic Docto?s Name};We are in search of volunteers for a study to examine the biomechanics of gait in patientswith medial osteoarthritis and using a Generation II Unloader brace. We are asking you toassist us in the search by bringing this to the attention of your patients. This study will bedone by Dr. David Sanderson and Peter Davidson at the Biomechanics Lab at the UniversityofBritish Columbia.The purpose of the study is to measure the effect the brace has on the varus angle (coronalangle) of the knee throughout the walking phase. Comparison will be made between no-braceand brace conditions. This research follows the study done by Horlick and Loomer (copyincluded) which found that the brace provides significant pain reliefDetails of the gait study experimental procedure and equipment is included in the attachedconsent form. We would appreciate you giving this to all patients that would meet the criteriabelow.Each volunteer should have unilateral medial compartment gonarthrosis (OA) and is using aGeneration II Unloader brace with a medial hinge as conservative treatment. The patientshould have owned their brace for a minimum of one month and are using it on a daily basisThe complete inclusion and exclusion criteria are listed on the next page. Please note thatthe subjects will be asked to walk briskly on a treadmill. It would not be appropriate torecruit a subject with heart difficulties or one on any medication that affects the cardiovascular response.If you have any questions please be free to call Dr. Sanderson, Mr. Davidson or Dr. Loomer.52SUBJECT RECRUITMENT CRITERIAInclusion Criteria:1) Male, 40-65 years old.2) Medial OA in the left knee only.3) Has owned a Medial Hinge Gil Unloader brace more than one month andwears it on a daily basis.4) Has pain while walking without brace.5) Has pain relief with brace.6) Typical muscle build, not too athletic.7) Manual dexterity adequate for self application of brace.Exclusion criteria:1) Arthritides other than OA.2) Previous fracture of ipslateral femur or tibia.3) Previous surgery to the affected knee other than arthroscopy, debridement,or partial menisectomy.4) Fixed flexion deformity greater than 15 degrees.5) Flexion less than 115 degrees.6) Leg length discrepancy greater than 2 centimeters.7) Skin disease or peripheral vascular disease preventing brace application.53SUBJECT INFORMATION FORMThe University ofBritish ColumbiaBiomechanics LaboratoryTitle of Investigation: THE EFFECTS OF VALGUS BRACING ON THE 3DKINEMATICS OF GAIT iN PATIENTS WITH OSTEOARTHRITIS OF THE KNEE.The purpose of this study is to provide a measure of the effect the Generation II knee bracehas on the walking pattern of people with single-joint arthritis. The idea is that the bracetends to straighten the leg when your foot is on the ground. This new study will contribute inthe understanding of how arthritis can be treated non-surgically and without drugs.Procedures: You will be asked to walk briskly on a treadmill while wearing small reflectivemarkers and foot contact indicators. Two cameras will videotape your movements. You willbe filmed several times by the video cameras while you are wearing or not wearing a kneebrace. After you are done, the computer will calculate the movements of the markers onyour legs. Your total time in the lab should not exceed one hour in length.For the setup, six markers will be placed on one leg with removable tape. One marker will beplaced on the side of your hip and on the outer aspect of your knee, ankle and foot. Finallyone marker will be placed on the front of your thigh and on the front of your lower leg. Eachmarker is a small reflective ball about the same size as a ping pong ball. We will providerunning shoes. Two foot contact indicators will be taped to the bottom of each shoe. Theindicators are small flat disks and you should not feel them when you are walking. Thin wiresfrom the indicators will be taped to follow up your leg and across to the computer. The wireswill not restrict your motion. The treadmill is a simple belt driven device. If needed, the beltcan be stopped at a moments notice. The frontal area of the treadmill is surrounded by safetybars. An attendant will be standing behind you on the platform. Throughout the session thebelt will never go faster than normal walking speed. At first the speed will be set as low asyou want until you are use to walking on the belt.During the session one video camera will be positioned far in front of you and the other off toyour side. The cameras will only film you for about two minutes each trial, One set of trialswill be the no-brace condition: ie you will not be wearing your brace. The computer will usethe videotape recordings to measure the angle ofyour leg at your knee. It is of interest to seehow this angle changes as you walk. During the other set of trials you will be wearing yourbrace. The procedure is the same as the no-brace condition. The brace will not be altered orchanged for this study. At a later date the computer will use the data collected to construct astick figure of your walking. The research will be based on comparing the no-brace conditionto the brace-condition.This study is expected to cause no harm or discomfort to you. Your identity will beconcealed and your name will never be used on any report or paper. You will have theopportunity to ask questions throughout the testing.54In signing this consent form you acknowledge that you have read and understood thedescription of the experiment you will be a part of You acknowledge that you are willinglyentering the experimental procedures and understand that you may withdraw at any timewithout consequence. You acknowledge that all the information about you will be kept inconfidence.CONSENT: I HAVE READ THE ABOVE COMMENTS AND UNDERSTAND TUEEXPLANATION WHICH IS GWEN AND I WILLiNGLY ENTER INTO THETESTING PROCEDURES.Subject’s signature DateWitness’s signature DateInvestigator’s signature Date55APPENDLX BTHIGH AXIAL ANGLE AN]) SHANK AXIAL ANGLEVECTOR FORMULAS56THIGH AXIALAXIAL VIEWANGLECALCULATION DSIIIII- LB_C-B EtJz..____xa0-rI —SACtflTAL VIEW CORONAL VIEW- -knee-hip vector: A reference vector: Eknee-thigh vector: thigh midpoint angle: All-parallel vector: C thigh axial angle: DE-axial vector: D7 oW 7I—CDCDEC-:©•eCD©nrJ2_a CaPfl[.II—II)CDI>—II—r-tnn-Ct—.II+—© Crc_s-+t•_I_at)+OaII+++II I>a+rI-L—lCD+-I--2>C-) II+>w —vI-fl——a58axial vector: D = D1i +D2j +D3knote.. D=B-Cthus.. D1=B—C , D2B—C B3- C3Dl =reference vector: f= .E2j +E3kwhere.. E2=A3 , E= -A2 , IEI =thigh axial angle: cos(i5 (D1E+2D3E)IDIIEI-tD/SHANKAXIAL_--ANGLECALCULATION JB eec-- nvector formulassame as thigh angle-e59APPENDIX CANALYSIS OF VARIANCE DATANUMBER OF CASES PROCESSED: 12DEPENDENT VARIABLE MEANS AND STANDARD DEVIATIONS ARE TNDEGREESDEPENDENT VARIABLESS1B1 : SPEED 1, NO-BRACE CONDITIONS1B2: SPEED 1, BRACE CONDITIONS2B1 : SPEED 2, NO-BRACE CONDITIONS2B2: SPEED 2, BRACE CONDITIONREPEATED MEASURES FACTORS AND LEVELSDEPENDENT VARIABLESWITHIN FACTOR 1 2 3 4SPEED 1 1 2 2BRACE 1 2 1 2SIGNIFICANT P VALUES (P<O.05) ARE IN BOLD FACE60FLEXION ANGLE. hEEL STRIKEDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B13.586(2.90)S 1B25.386(3.79)S2B 12.030(3.28)UIVARIATE REPEATED MEASURES ANALYSISS2B22.468(4.07)SOURCE Ss DF MS F PSPEED 60.036 1 60.036 17.135 0.002ERROR 38.541 11 3.504BRACE 15.023 1 15.023 3.909 0.074ERROR 42.277 11 3.843SPEED*BRACE 5.562 1 5.562 2.352 0.153ERROR 26.016 11 2.365FLEXION ANGLE. MAX FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B212.015 12.838 14.504 15.699(4.46) (4.79) (5.07) (5.46)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 85.856 1 85.856 23.806 0.000ERROR 39.672 11 3.607BRACE 12.214 1 12.214 4.622 0.055ERROR 29.066 11 2.642SPEED*BRACE 0.416 1 0.416 0.489 0.499ERROR 9.359 11 0.85161FLEXION ANGLE. MIN FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B12.411(4.17)S 1B24.809(4.63)S2B 12.120(4.73)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MSS2B24.800(4.98)FSPEED 0.272 1 0.272 0.097 0.761ERROR 30.839 11 2.804BRACE 77.374 1 77.374 23.042 0.001ERROR 36.938 11 3.358SPEED*BRACE 0.239 1 0.239 0.271 0.613ERROR 9.725 11 0.884FLEXION ANGLE TOE-OFFDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B232.266 31.068 31.160 30.965(7.61) (7.15) (9.58) (7.07)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 4.386 1 4.386 0.209 0.656ERROR 230.861 11 20.987BRACE 5.815 1 5.815 0.552 0.473ERROR 115.941 11 10.540SPEED*BRACE 3.020 1 3.020 0.347 0.568ERROR 95.746 11 8.704P62DEPENDENT VARIABLE MEANS (&S 1B2-0.914(1.92)UNIVARIATE REPEATED MEASURESSOURCE SS DFSTANDARD DEVIATIONS)S2B1 S2B2-0.305 -0.816(2.18) (2.18)ANALYSISMS FTHIGH CORONAL ANGLE. HEEL STRIKES1B1-0.713(2.17)PSPEED 0.767 1 0.767 0.480 0.503ERROR 17.572 11 1.597BRACE 1.520 1 1.520 1.748 0.213ERROR 9.566 11 0.870SPEED*BRACE 0.290 1 0.290 1.667 0.223ERROR 1.916 11 0.174THIGH CORONAL ANGLE. MAX FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B2-1.201 -1.102 -0.521 -0.739(1.81) (1.78) (1.79) (1.74)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 3.266 1 3.266 4.263 0.063ERROR 8.428 11 0.766BRACE 0.043 1 0.043 0.062 0.808ERROR 7.566 11 0.688SPEED*BRACE 0.300 1 0.300 1.275 0.283ERROR 2.588 11 0.23563THIGH CORONAL ANGLE. MIN FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1-2.783(1.78)S 1B2-2.399(1.56)S2B1-2.111(1.81)UNIVARIATE REPEATED MEASURES ANALYSISS2B2-1.905(1.85)SOURCE SS DF MS F PSPEED 4.082 1 4.082 12.662 0.004ERROR 3.546 11 0.322BRACE 1.042 1 1.042 1.705 0.218ERROR 6.722 11 0.611SPEED*BRACE 0.095 1 0.095 0.590 0.459ERROR 1.763 11 0.160THIGH CORONAL ANGLE. TOE OFFDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 51B2 S2B1 S2B21.832 1.827 1.646 1.745(1.72) (1.83) (1.46) (1.37)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 0.216 1 0.216 0.433 0.524ERROR 5.496 11 0.500BRACE 0.026 1 0.026 0.079 0.784ERROR 3.613 11 0.328SPEED*BRACE 0.033 1 0.033 0.212 0.654ERROR 1.690 11 0.15464ShANK CORONAL ANGLE. HEEL STRIKEDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S lB 11.882(1.99)S 1B21.446(1.66)S2B 11.249(1.81)UNIVARIATE REPEATED MEASURES ANALYSISS2B21.193(2.01)SOURCE SS DF MS F PSPEED 2.350 1 2.350 6.23 1 0.030ERROR 4.149 11 0.377BRACE 0.725 1 0.725 0.917 0.359ERROR 8.698 11 0.791SPEED*BRACE 0.434 1 0.434 0.801 0.390ERROR 5.966 11 0.542SHANK CORONAL ANGLE. MAX FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B24.265 4.026 3.720 3.637(2.27) (1.99) (2.13) (2.30)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 2.621 1 2.621 3.606 0.084ERROR 7.996 11 0.727BRACE 0.3 12 1 0.3 12 0.575 0.464ERROR 5.958 11 0.542SPEED*BRACE 0.072 1 0.072 0.238 0.635ERROR 3.342 11 0.30465SHANK CORONAL ANGLE. MIN FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B13.507(1.98)S 1B23.201(1.87)S2B 13.254(2.06)UNIVARIATE REPEATED MEASURES ANALYSISS2B23.086(1.87)SOURCE SS DF MS F PSPEED 0.407 1 0.407 1.291 0.280ERROR 3.473 11 0.3 16BRACE 0.672 1 0.672 1.053 0.327ERROR 7.015 11 0.638SPEED*BRACE 0.058 1 0.058 0.167 0.691ERROR 3.821 11 0.347SHANK CORONAL ANGLE. TOE-OFFDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B26.619 5.310 7.387 6.951(3.51) (3.08) (3.91) (4.44)lINTVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 17.423 1 17.423 5.223 0.043ERROR 36.694 11 3.336BRACE 9.130 1 9.130 6.804 0.024ERROR 14.761 11 1.342SPEED*BRACE 2.286 1 2.286 0.927 0.356ERROR 27.134 11 2.46766THIGH AXIAL ANGLE. HEEL STRIKEDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1-10.748(3.25)S 1B20.345(6.15)S2B 1-16.314(4.86)UNIVARIATE REPEATED MEASURES ANALYSISS2B2-2.7 19(6.42)SOURCE SS DF MS F PSPEED 223 .434 1 223 .434 74.995 0.000ERROR 32.773 11 2.979BRACE 1828.463 1 1828.463 62.337 0.000ERROR 322.652 11 29.332SPEED*BRACE 18.785 1 18.785 21.283 0.001ERROR 9.709 11 0.883THIGH AXIAL ANGLE. MAX FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B2-3.551 4.458 -5.951 3.986(3.06) (6.00) (2.83) (4.84)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 24.758 1 24.758 11.332 0.006ERROR 24.033 11 2.185BRACE 966.196 1 966.196 59.614 0.000ERROR 178.282 11 16.207SPEED*BRACE 11.145 1 11.145 5.330 0.041ERROR 23.002 11 2.09167THIGH AXIAL ANGLE. MIN FLEXDEPENDENT VARIABLE MEANS (&S1B1 S1B2-0.448 4.274(3.42) (5.83)UNIVARIATE REPEATED MEASURESSOURCE SS DFSTANDARD DEVIATIONS)S2B1 S2B2-0.551 5.788(4.31) (5.82)ANALYSISMS FTHIGH AXIAL ANGLE. TOE-OFFDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B25.724 9.757 6.253 10.882(3.45) (6.15) (6.25) (4.96)UMVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 8.202 1 8.202 1.987 0.186ERROR 45.415 11 4.129BRACE 225.133 1 225.133 15.789 0.002ERROR 156.849 11 14.259SPEED*BRACE 1.069 1 1.069 0.584 0.461ERROR 20.146 11 1.831PSPEED 5.975 1 5.975 3.863 0.075ERROR 17.015 11 1.547BRACE 367.049 1 367.049 23.816 0.000ERROR 169.533 11 15.412SPEED*BRACE 7.857 1 7.857 4.591 0.055ERROR 18.828 11 1.71268SHANK AXIAL ANGLE, REEL STRIKEDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B2-1.803 -4.534 -2.040 -4.820(3.42) (3.56) (3.98) (3.09)UNTVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 0.821 1 0.821 0.823 0.384ERROR 10.982 11 0.998BRACE 91.098 1 91.098 8.925 0.012ERROR 112.275 11 10.207SPEED*BRACE 0.007 1 0.007 0.009 0.926ERROR 8.858 11 0.805SHANK AXIAL ANGLE. MAX FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 52B1 52B2-1.487 -0.220 -2.189 -0.027(3.99) (4.15) (3.30) (3.09)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 0.779 1 0.779 0.250 0.627ERROR 34.343 11 3.122BRACE 35.268 1 35.268 4.449 0.059ERROR 87.199 11 7.927SPEED*BRACE 2.397 1 2.397 4.333 0.062ERROR 6.085 11 0.55369ShANK AXIAL ANGLE. MIN FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B2-0.529 -3.470 0.695 -1.824(3.80) (4.80) (4.13) (5.55)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 24.705 1 24.705 9.98 1 0.009ERROR 27.228 11 2.475BRACE 89.386 1 89.386 5.112 0.045ERROR 192.343 11 17.486SPEED*BRACE 0.533 1 0.533 0.772 0.398ERROR 7.596 11 0.691SHANK AXIAL ANGLE. TOE-OFFDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)S1B1 S1B2 S2B1 S2B27.227 9.778 10.446 13.594(5.07) (6.49) (5.56) (6.70)UMVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 148.504 1 148.504 11.255 0.006ERROR 145.135 11 13.194BRACE 97.411 1 97.411 8.294 0.015ERROR 129.199 11 11.745SPEED*BRACE 1.069 1 1.069 0.157 0.699ERROR 74.834 11 6.80370EVENT 2 TIMING, MAX FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)units in percent of step cycleS1B1 S1B2 S2B1 S2B213.333 13.500 15.000 15.833(2.21) (2.02) (3.22) (3.31)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 48.000 1 48.000 13.895 0.003ERROR 38.000 11 3.455BRACE 3.000 1 3.000 1.737 0.214ERROR 19.000 11 1.727SPEED*BRACE 1.333 1 1.333 1.692 0.220ERROR 8.667 11 0.788EVENT 3 TIMING, MIN FLEXDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)units in percent of step cycleS1B1 51B2 S2B1 S2B241.500 41.167 38.833 38.500(2.72) (4.12) (2.76) (3.18)UNIVARIATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 85.333 1 85.333 32.744 0.000ERROR 28.667 11 2.606BRACE 1.333 1 1.333 0.122 0.734ERROR 120.667 11 10.970SPEED*BRACE 0.000 1 0.000 0.000 1.000ERROR 54.000 11 4.90971EVENT 4 TIM1NG TOE-OFFDEPENDENT VARIABLE MEANS (& STANDARD DEVIATIONS)units in percent of step cycleS1B1 S1B2 S2B1 S2B261.230 61.824 58.694 58.701(2.68) (2.92) (2.27) (2.35)UNIVARJATE REPEATED MEASURES ANALYSISSOURCE SS DF MS F PSPEED 96.059 1 96.059 53.576 0.000ERROR 19.722 11 1.793BRACE 1.084 1 1.084 0.570 0.466ERROR 20.909 11 1.901SPEED*BRACE 1.034 1 1.034 1.128 0.311ERROR 10.085 11 0.917APPENDIX DINDIVIDUAL SUBJECT AND GROUP AVERAGEANGLE GRAPHS72CCICIIc_mmimC,C,IC,-I-II_IiItSSSSSSSSSSS00CICCc_IcmlmmC)C)rnrn-F’Irn22C)rn00CCwCmmC)C)--Il%3-oLiELI0mC-rn2C)Irn2C)m0CG10a.CSSS*SSSC)0CmC)m0CCmC)-INCI-wzmrn74C,C’,IC.)LU-DU)NLUU)S S S * S S S SLULUC)0zLUC)0.NI-C)LU-)Cl)0NI-C)LUDU)LU-JC,zz0xLU-JU-LULUzSSSS.SCC)z1%I-C)LU-)0CDI-C)LU-,D0CDI-C)LU-)0I-C)LU-)D0-I-NI-C)LU-)zC,)II-C)LUD00I-C)LUD0LUC,LU>0.0C,0)I-C)LU0I—0)I-C)LU0)0LUC,zLU-JC,zL, Sb4) C4) C5) 1).U •o 0r- —LULU00 0-JS S * * 5)2 a S S * S S 2C0-v<•0mzC)Ioo0.0.COG0COCOCOC)0C•0C’)CCoCm0-I0C-Co0CCoCmC-I0CCoCm0-INCI-0CCoCm0CO-I0CCoCmC)-ICoC0CCoCm0-IN0Co0CCoCm0-IN-v0CCoCmC)-I0CCoCmC)-lC,’0CCoCm0-ICo0CCoCmC)-I2C)-IC)0002I2C)ImCo20C’)F”Fil0CCoCm0-IC)cLI0mCmzC)I-00..z ••C)m•• ••00C)0CmC)m0CwCmC)-I0-I0C0mC)0C0C0mC)-IN000C0Crn1C)HN-V0C0CrnC)-IC’,C0CrnC)-IU’0C0CrnC)-‘0,0C0CmC)-lzC)0C0CrnC)-I0C-00C0CrnC)-IC-I0C0CrnC)-INCI-0C0CrnC)-IC)-IC)=C)002I2C)Irn0rnm0NO2C)0rn.59L77COIC.)LUCoLUUi0.CoLUozLU-I0z-Jz00C.)zCo2C.)I..IC.)LU-)CoIC.)LUCoI-C.)LU-)D0IC.)LU-,D00.NIC.)LU00NI.C.)LUD0D0)IC.)Ui0I0)IC)LU0)D0-jNIC.)LU0)CoI-,IC.)LU-,0)DCl)0)-)CC.)Ui0)CoiCCLUC,LU>0.0C,2 •C CC VLUV Vz • •00LU-ICDLULU0.C. 0z -JrC0<mzC)I—0ooD..•••G00CCDCmC)-I0CCDCmC)H‘010CCDCmC)-I0C)0C•0mC)m0CCDCmC)-I0CCD0CCDCmC)-C-I0CCDCmC)-INCI-0CCDCmC)-ICD-l0CCDCmC)-ICDC0CCDCmC)-IN0CD0CCDCmC)-IN•002C)002IzC)ImCD2CDm0CCDCmC)-I2C)0-vmmN0CCDCmC)-l08Lo•-V20Imoo--z mmmmm mm00SSSSSS-S ‘S00CC00CC-4-IS11iI00C0CCmC)-I0C0CCmC-I0CmCI--4IC)IIz6)Im/2C,C,m-C0rnrnCnCCmC,-4(‘3C)6L‘ComzC)Im000.0.C)m0000Cl,CCmC)-I-*0CCmC)-I0C0CCm0-I/C)0CmC)m0CCmC)-f0CCmC)-IU’0CCm0-IU,0CU,CCli0-I20CS.5C..5-IC)xI2C)Im0)CU,CCli0-IMaU,0CU,CmC)-I•0U,C)U,0Im0mmIi.)0CU,Cm0-lCS,C)088.8I-C0-Q-vmzC)caiii00Q.Q-zC)mm••CCmC)-I0Cw0CmC)m0CCmC)-IC-I0CCmC)-I-I0CCmC)-ICoC0CCmC,-IN0w0CCmC)-IN-Q-Q0CCmC)-I0CCmC)-lU’0CCmC)-I0CCmC)-IzC).80CCmC)-INCI-0mm0CCmC)-l(5)C)18rC00.cm2C)rooaaO0o0aaaa00(0CwCmC)-I0CwCmC.)-la’0CwCmC)-Ia’C)0CmC)m0Ca’C‘iiC,-I(0-I0Ca’C..mC)-I(0C0Ca’CItC,-lN0a’0Ca’CItC)-IN•0-C0x2Pcr2C)rm0Ca’CItC,-I0Ca’0Ca’CItC,-IC--l0Ca’CItC,-INCI0Ca’CItC,-IC,)C)‘SC’S‘SSSIa’2C)ItCC!CItC,-I2C)0It0N


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items