Impact biomechanics of lateral knee bracing The anterior cruciate

ligament

LONNIE E. PAULOS,* MD, PATRICK W.

CAWLEY,† OPA, RT, E. PAUL FRANCE,‡§

AND PhD

From *Salt Lake Knee and Sports Medicine, and the ‡Orthopedic Biomechanics Institute, Salt Lake City, Utah, and the †DonJoy Biomechanics Research Laboratory, Carlsbad, California

ABSTRACT

models, have shown that the level of medial collateral ligament (MCL) protection given by even the best braces was inconsistent and only marginap,5 In all of these studies, the effect of these braces on the ACL has been given little

We evaluated the effects of six different prophylactic braces on ACL ligament strain under dynamic valgus loads using a mechanical surrogate limb validated against human cadaveric specimens. Medial collateral ligament and anterior cruciate ligament peak forces, medial collateral ligament and anterior cruciate ligament tension initiation times, and impact safety factors were calculated for both braced and unbraced conditions. These tests were conducted to determine whether or not application of a prophylactic brace might provide protection to the anterior cruciate ligament under valgus loading conditions. The results of this study indicate that those braces that increased impact duration appear to differentially protect the anterior cruciate ligament more than the medial collateral ligament, and that most of the braces tested appear to provide some degree of protection to the anterior cruciate ligament under direct lateral impacts. These findings should be confirmed clinically.

attention. This study was initiated to determine the effects of the use of prophylactic braces on the ACL under direct dynamic valgus loading of the knee.

MATERIALS AND METHODS The mechanical surrogate developed by France et a1.2 for the evaluation of knee/brace valgus impact characteristics was used in this investigation. This mechanical surrogate is an anatomically emulated knee model instrumented for measurement of knee ligament tension as a function of medial joint opening. The surrogate was validated against human cadaveric data and its dynamic loaded response was within 2% of the normalized cadaveric data for direct valgus loads. The ACL and MCL are modeled using teflon-coated, braided stainless steel cables in series with compression springs and load cells. Tibial and femoral attachment points were chosen to mimic in vitro strain patterns obtained from valgus testing of cadaveric specimens. This surrogate incorporates a contralateral limb, weighted torso, and functional hip and ankle units on the instrumented side (Fig. 1). This specimen can be tested completely unconstrained as well as constrained, with the torso, pelvis, hip, and foot on the instrumented side rigidly fixed. For this study, only the constrained condition was used, as it was considered to represent the &dquo;worst case&dquo; scenario. A soft tissue covering with the approximate compliance and contour of a normal human limb was placed over the instrumented limb to permit accurate brace placement. This surrogate closely duplicated the dynamic response of the cadaveric model to direct valgus loads, but its applicability to the in vivo human model requires further clinical validation. A custom-designed trolly system was used to deliver valgus

Despite a good deal of recent investigation, the effectiveness of lightweight lateral knee braces in preventing valgusrelated knee injuries is still a matter of controversy among health care providers. While circumstantial and anecdotal testimony tends to support the use of lateral bracing, the results of both clinical and biomechanical studies have been

equivocal. Clinical studies on this subject have been retrospective or short term and show conflicting results even when the same athletic teams are studied during the same time frame.’,’ 3 4,6,8,10 Biomechanical studies on lateral bracing, conducted on either cadaveric or mechanical surrogate § Address correspondence and repnnt requests to E Paul France, PhD, Intermountain Orthopedic Research Laboratory, 359 8th Avenue, Suite 206, Salt Lake City, UT 84103 337

338

ligament tension, peak impact load, and trolley velocity were recorded as functions of time using a computer based analog/ digital acquisition system with a 50 Hz sampling rate. Data acquisition was triggered by the trolley at the same point relative to the lateral aspect of the knee for each test. Impact tests were accomplished with the surrogate knee in either 0° or 30° of flexion with the ankle and foot fixed medially (rotation and vertical motion allowed), and with a fully weighted torso. Individual test conditions are listed below.

For each test condition the unbraced knee was first impacted three times. A brace was then applied and three additional impacts were delivered. This same procedure was repeated for each brace type. Only direct lateral blows at the level of the joint line were used in this study. The order of brace testing was randomized for each test condition. A total of 72 braced and 72 unbraced tests were performed. The braces tested were: Omni-Anderson Knee Stabler (Omni Scientific, Inc., Lafayette, IL), McDavid Knee Guard (McDavid Knee

Guard, Inc., Clarendon Hills, IL), Stromgren (StromgrenScott, Inc., Hays, KS), SMI PKG (Sports Medicine, Inc., Carlsbad, CA), Mueller (Mueller Sports Medicine, Inc., Prairie Du Sac, WI), and TruFit Renegade (Tru-Fit, Inc., Lynn, Figure

1. The

surrogate knee used

in our

Figure 2. Impact test configuration (right) in position for lateral impact.

MA).

investigation.

showing impact trolley

Several parameters were analyzed to describe brace performance. The time difference between the onset of braced and unbraced ligament tension was measured to determine whether application of a lateral brace affects the onset of ligament tension. Peak ligament loads for the ACL and MCL in braced versus unbraced tests were also compared. Data for individual tests was compared using the Student’s t-test to determine statistical differences. The impact safety factor (ISF), described in our previous work,2was also calculated for both the MCL and ACL for each test condition. The ISF was calculated using the following equation:

For example, an ISF of 1.50 would represent a 30% reduction in ligament force for the same impact conditions unbraced. Calculation of the ISF facilitates brace-to-brace comparisons.

instrumented knee (Fig. 2). The system consisted of a ramp and gravity-accelerated impact trolley. The mass of the trolley could be varied from 50 to 300 pounds (22.68 to 136.08 kg), and was capable of velocities up to 9 mph, (14.49 km/hr). For each test, ACL and MCL

impact loads

to the

All tests were videotaped to determine the presence of joint line contact of the brace at impact and to determine positional changes of the orthotic devices with impact. All braces were inspected following each impact to determine if permanent deformation or visible damage was present. De-

339

was determined by a pretest measurement of the cuff-to-hinge bar distance while the brace rested with its

formation

limb side on a flat surface. Each brace was measured following each test condition.

RESULTS None of the braces in this investigation sustained significant permanent deformation greater than 1 %. This slight amount

of deformation could have been the result of a flattening of the attachment cuff while in contact with the limb or may have been due to a slight bending of the hinge bar itself. While every effort was made to ensure that braces were aligned directly on the impact axis, those braces with less surface-limb contact areas tended to rotate a small amount anteriorly with valgus impact. In the more compliant braces, contact of the brace at the lateral joint line occurred with every impact. Joint line contact occurred in the less compliant braces only with the higher impact momentum. Those braces that permitted joint contact at all impact loads were the least effective at reducing peak loads in both the ACL and MCL. There was no increase in passive MCL strain with the application of any brace. For almost all braces, braced ACL tension initiation times were either the same or longer than those for the unbraced condition (Table 1). This indicates that the majority of braces tended to lessen the impact momentum resulting in an increased impact duration under most test conditions. This would also indicate that velocity of valgus joint opening was lessened as well. There was only a minor decrease in this time span at the 30°, 170 pounds, 2.75 mph test condition in one of the braced specimens. As shown in Table 1, the time to ACL tension initiation for most braces was greater at 0° of knee flexion than at 30° of flexion. The P values listed in Table 1 compare braced versus unbraced tension initiation times.

An evaluation of the ratio of ACL to MCL peak force, as well as the mean ACL and MCL peak forces, indicated that, in most cases, application of a prophylactic brace, regardless of design, resulted in a larger decrease in ACL peak force as compared to the decrease in MCL peak force. The individual mean data for percent reduction of MCL peak load and MCL ISF are presented in Table 2. Although there was a general trend toward reduction in MCL peak forces for all braces, individual behavior varied and was affected by test condition. The mean reduction in MCL peak load was 21.95% ± 6.92% and the range was 10.02% to 41.20% for all tests. The mean ISF for the MCL was 1.29 ± 0.12, with a range of 1.12 to 1.70 for all tests. In our previous study,2 we proposed that an ISF of 1.50 be the minimal accepted standard. When ISF data was compared statistically, the SMI PKG proved to be statistically superior to the following braces at the following P values: McDavid (0.005), Stromgren (0.03), TruFit (0.01), and Mueller (0.007). There was no statistical difference between the SMI PKG and the Omni-Anderson and there were no statistical differences when comparing all other braces to one another. The individual mean data for percent reduction in ACL peak load and ISF are shown in Table 3. As with the MCL, individual brace performance varied with the test condition; however, the reductions in ACL peak loads were proportionately larger for all braces when compared to the MCL data. The mean reduction in ACL peak load was 38.9% ± 15.32%, with a range of 8.05% to 74.5% for all test conditions. The mean ACL ISF was 1.78 ± 0.60, with a range of 1.09 to 3.92 for all test conditions. Comparisons of mean percent reduction in ACL versus MCL peak loads are presented graphically in Figure 3. ISFs for both the ACL and MCL are compared graphically in Figure 4. When ACL mean ISF data was compared statistically, the SMI PKG again demonstrated significant improvement when compared to the P values of the following braces: McDavid Knee Guard (0.004),

TABLE 1 MCL-ACL tension initiation time difference (unbraced minus braced)’

° *, denotes significance less than 0.05. Data in parentheses are standard error of the difference. b McD, McDavid Knee Guard; STROM, Stromgren; OMNI, Omni-Anderson Knee Stabilizer; PKG, SMI PKG; MUEL, Mueller; TRU, TruFit.

340

TABLE 2 MCL peak load&dquo;

° Percent reduction (%RED) and ISF. 6 See definitions at Table 1.

Figure 4. Mean ISF for ACL and MCL.

TABLE 3 ACL peak load

° Percent reduction (%RED) and ISF. b See definitions at Table 1.

Figure 3. Mean percent load braced) of the MCL and ACL.

reduction

(braced

versus un-

Stromgren (0.03), Omni-Anderson (0.04), and (0.005). There

was no

Mueller

significant difference between the

SMI PKG and TruFit. The TruFit also demonstrated a statistical improvement over the McDavid Knee Guard at the 0.05 level.

DISCUSSION This

investigation

was

conducted

using

rogate of the human lower limb that

a

was

mechanical

sur-

developed from,

and validated against, dynamic valgus failure data for human cadaveric specimens. The dynamic valgus response of this mechanical surrogate is within 2% of the normalized cadaveric failure data for all test conditions. Thus it can be said that this surrogate model closely mimics the cadaveric model for dynamic valgus failure. Whether or not this data is applicable to the in vivo model remains an open question that can only be answered through additional carefully controlled clinical evaluations. We leave it to the reader to assess the significance of the data presented here. Once applied to the limb, the brace and limb form a composite structure whose response is ultimately dependent upon the material properties of both structures. Since the principal interface between brace and limb is soft tissue, it is certain that the factors affecting soft tissue compliance and response-such as friction, viscosity, elasticity, and inertia-will also affect the performance of a brace under load. For example, it is likely that the impact behavior of soft tissue is rate-dependent and that soft tissue compliance is critical in the distribution, absorption, and transmission of dynamic valgus loads. Unfortunately, little is known about soft tissue compliance. Further study will be necessary before a detailed description of composite brace-limb response will be possible. In a previous study,’ we identified three critical mechanical factors that determine lateral brace function under dynamic valgus loads: absorption, distribution, and transmission of energy. Absorption describes how the brace manages impact energy through deformation. Distribution primarily describes brace design in regard to the distribution of impact forces across the brace and knee as a composite structure. Transmission refers to the energy imparted to the braced structure, i.e., the knee, and is dependent upon both brace material properties and design. The brace that performed best for most test conditions in this study was constructed of a relatively stiff aluminum alloy and its design mimicked a &dquo;leaf spring&dquo; on an automobile. The material properties and design permitted deformation on initial impact, which tended to increase the impact duration and slow the rate of load. This design also prevented joint line contact except under the highest loading conditions. The length of this brace and broad contact areas on the soft tissues of the

341

thigh and calf helped to distribute loads away from the joint and may have taken advantage of soft tissue damping effects. It should be noted that braces of a similar design, but constructed of resilient plastic, also performed well in reducing ACL peak loads. Less resilient plastic designs permitted early joint line contact, allowing transmission of more energy directly to the joint and resulting in higher peak loads in both the ACL and PCL. As with the MCL, brace design, brace fit, and the direction of load all play a prominent role in the effectiveness of lateral knee bracing in protecting the ACL. Although MCL protection was variable and of questionable significance with current brace designs, most braces appear to offer significantly greater protection to the ACL under direct valgus loads. The reasons for this are not only due to the mechanical features mentioned above, but may also be due to the lateral position of the brace relative to the ACL versus brace position relative to the MCL. The increase in ACL tension initiation times was bracedependent and appears to be the result of an increase in impact duration. This phenomenon was related to both brace design and material properties that determine energy absorption characteristics. Increased brace stiffness to bending seems to be an important factor, but must also be coupled with the geometry of the brace. Length of the brace, standoff from the joint line, and brace/leg fit also play important roles. The actual material used in brace construction seems to be less important than its dynamic deformable behavior since braces constructed of both plastic and metal gave the best results in this study. The same features also combine to delay the time, not only for ACL tension initiation, but also for joint line contact, which could create a three-point bending system and reduce brace effectiveness. The mechanical interface between the brace and limb also appears to be an integral factor in determining the function of a prophylactic brace. The viscoelastic nature of the soft tissues contributes to brace performance in several ways. When a dynamic valgus load is applied, the brace is compressed toward the limb. The lack of resistance to tangential loads in the soft tissues permit axial motion of the brace bars and, thus, allows the brace to deform and absorb some of the load as well as change the rate of loading. The viscoelasticity of the tissues further aids absorption of energy through dispersal into the tissues and provides a damping effect, reducing the acceleration effects transmitted to the joint. Braces with larger bearing surfaces take greater advantage of this phenomenon. For most of the braces tested, ACL tension initiation time was greater with the knee at 0° of flexion than when the knee was flexed to 30°. The same trend is true for both MCL and ACL peak load data where the reduction in peak load was greater with the knee in extension. The differences in brace design and material properties seemed to be less important when the knee was extended. It appears that laterally applied braces are more effective and offer the greatest stiffness when axially aligned. As flexion increases, the laterally applied bars offer decreasing resistance to val-

bending. With the limb in extension, we noted little medial rotation of the knee with valgus impact. However, with 30° of flexion, we noted considerable medial rotation of the limb with valgus impact. This medial rotation may contribute to the increased MCL and ACL loads seen with flexion. This point serves to remind us that this study was conducted under carefully controlled conditions in which all impacts were direct valgus blows at the level of the joint line. How often this mechanism actually occurs on the field of play has yet to be determined. Therefore, one must use reasonable caution when extrapolating the results of this study to the actual sporting situation. It is our opinion that certain sports, and certain positions within those sports, are definitely vulnerable to direct lateral blows at the knee and, in well-controlled clinical studies, these sports and specific positions may well show some benefit from lateral prophylactic braces. A recent study3 determined that the use of a prophylactic knee brace might increase the risk of incurring injuries elsewhere in the lower limb complex, particularly in the ankle. In its current configuration, the mechanical surrogate does not allow for measurement of forces and moments at the foot and ankle. However, the authors of a recent wellcontrolled clinical study9 found no difference in ankle injuries among braced and unbraced players in a study using only one brace design. This may be a factor that is bracedependent and bears further investigation. In interpreting this data, the reader must be cognizant that there are both benefits and limitations in the use of a mechanical surrogate. A benefit of the mechanical surrogate is that it permits true parametric testing in that the response of the surrogate can be calibrated so that it is the same from brace to brace and condition to condition. While this system closely duplicates the response of the cadaveric model, it is only validated for valgus impact loads. Neither the ligament analogs nor the external soft tissue model are truly physiologic in their response to dynamic load, although they are good approximations based on cadaveric testing. This system does not permit conclusions regarding brace behavior other than those for direct lateral impacts. In our opinion, the mechanical surrogate is a valuable tool in the evaluation and comparison of brace performance and loading conditions. gus

CONCLUSIONS

study is one in a series of studies conducted at our laboratory over the past several years. Funding was obtained from several different sources including two different brace manufacturers, research foundations, and private contribuThis

tions. The authors have not received any direct benefit from the sale or manufacture of lateral protective braces. Our conclusions are based solely on our research data as well as a review of other publications to date. It is our opinion, based on this research and review, that the following conclusions can be made concerning prophylactic knee braces: 1. Lateral prophylactic knee bracing appears to have a

342

beneficial effect in protecting both the MCL and the ACL against direct lateral blows to the knee. This effect appears to be greater for the ACL than it is for the MCL. 2. Brace hinge contact with the lateral joint line of the knee reduces the effectiveness of lateral prophylactic bracing. This phenomenon is design-dependent and variable from brace to brace. 3. Although most presently available brace designs appear to provide some protection to the ACL and MCL under direct valgus loading conditions, there is a definite need for improved brace designs that we believe are both necessary and a reasonable possibility. 4. It is our opinion, based on the current research, that prophylactic knee bracing is a viable concept and that, at the present time, the use of presently available prophylactic knee brace designs should not be discouraged for contact sports and selected positions within those sports. The use of these braces should be voluntary and a system must be in place to ensure that they are properly applied and maintained in good condition at all times.

REFERENCES 1

Curran RD, Linquist DS: Statistical lactic knee braces Thesis, Duke

analysis of the effectiveness of prophyUniversity, Dept of Mechanical Engi-

neering,1986 2 France EP, Paulos LE, Jayarman G, et al The biomechanics of lateral knee bracing Part 2 Impact response of the braced knee Am J Sports Med 15 430-438, 1987 3 Grace TG, Skipper BJ, Newburry JC, et al Prophylactic knee braces and injury to the lower extremity. J Bone Joint Surg 70A 422-427, 1988 4. Hewson GF, Mendini RA, Wang JB. Prophylactic knee bracing in college football Am J Sports Med 14. 262-266, 1986 5 Impact testing of knee braces MKG and Anderson Stabler Models. McDavid Knee Guard, Inc Test Report 1/1984-1/1985. Gaynes Testing Laboratories, Inc , Clarendon Hills, IL 6 Moser KR: Analysis of the use of valgus knee braces in selected ntercollegiate football programs Thesis, Ohio University College of Education, 1985 7 Paulos LE, France EP, Rosenberg TD, et al The biomechanics of lateral knee bracing Part 1 Response of the valgus restraints to loading Am J Sports Med 15 419-429, 1987 8 Rovere GD, Haupt HH, Yates CS: Prophylactic knee bracing in college football Am J Sports Med 15 111-116, 1987 9. Sitler M, Ryan JB, Wheeler JH, et al The efficacy of a prophylactic knee brace to reduce knee Injuries in football Trans Orthop Res Soc 15 224, 1990 10 Teitz CC, Hermanson BA, Kronmal RA, et al Evaluation of the use of braces to prevent injury to the knee in collegiate football players. J Bone Joint Surg 69A: 2-9, 1987