You are on page 1of 19

Research Report

Agility and Perturbation Training
Techniques in Exercise Therapy for
Reducing Pain and Improving Function
in People With Knee Osteoarthritis:
G.K. Fitzgerald, PT, PhD, FAPTA, is A Randomized Clinical Trial
Associate Professor, Department of
Physical Therapy, School of Health G. Kelley Fitzgerald, Sara R. Piva, Alexandra B. Gil, Stephen R. Wisniewski,
and Rehabilitation Sciences, and Chester V. Oddis, James J. Irrgang
Director, Physical Therapy Clinical
and Translational Research Center,
University of Pittsburgh, 6035 Forbes Background. Impairment-based exercise programs have yielded only small to
Tower, Pittsburgh, PA 15101 (USA).
moderate benefits in reducing pain and improving function in people with knee
Address all correspondence to Dr
Fitzgerald at: kfitzger@pitt.edu. osteoarthritis (OA). It has previously been proposed that adding agility and pertur-
bation training to exercise programs for people with knee OA may improve treatment
S.R. Piva, PT, PhD, OCS, FAAMOPT,
effects for pain and function.
is Assistant Professor, Department
of Physical Therapy, School of
Health and Rehabilitation Sciences,
Objective. The purpose of this study was to examine the effectiveness of adding
University of Pittsburgh. agility and perturbation techniques to standard exercise therapy compared with the
standard exercise program alone for people with knee OA.
A.B. Gil, PT, PhD, is Research Spe-
cialist, Department of Physical Ther- Design. This was a single-blinded randomized controlled trial.
apy, University of Pittsburgh.

S.R. Wisniewski, PhD, is Professor, Setting. The study was conducted in the outpatient physical therapy clinic of a
Department of Epidemiology, and large, university-based health center.
Co-Director, Epidemiological Data
Center, Graduate School of Public Participants. One hundred eighty-three people with knee OA (122 women, 61
Health, University of Pittsburgh. men) participated.
C.V. Oddis, MD, is Professor of
Medicine and Director, Fellowship Interventions. Participants were randomly assigned to either a group that
Training Program, Division of received agility and perturbation training with standard exercise therapy or a group
Rheumatology and Clinical Immu- that received only the standard exercise program.
nology, University of Pittsburgh
School of Medicine, Pittsburgh, Measurements. The outcome measures were self-reported knee pain and func-
Pennsylvania. tion, self-reported knee instability, a performance-based measure of function, and
J.J. Irrgang, PT, PhD, ATC, FAPTA, global rating of change.
is Associate Professor and Director
of Clinical Research, Department of Results. Although both groups exhibited improvement in self-reported function
Orthopaedic Surgery, University of and in the global rating of change at the 2-, 6-, and 12-month follow-up periods, there
Pittsburgh School of Medicine.
were no differences between groups on these outcomes. There was no reduction in
[Fitzgerald GK, Piva SR, Gil AB, et al. knee pain or improvement in performance-based function in either group.
Agility and perturbation training
techniques in exercise therapy for Limitations. It is possible that more-intense application of the interventions or
reducing pain and improving func- application of the interventions to participants with knee OA who were at greater risk for
tion in people with knee osteo-
falling may have yielded additive effects of the agility and perturbation training approach.
arthritis: a randomized clinical trial.
Phys Ther. 2011;91:452– 469.]
Conclusions. Both intervention groups exhibited improvement in self-reported
© 2011 American Physical Therapy function and the global rating of change. Our results, however, did not support an
Association additive effect of agility and perturbation training with standard exercise therapy in our
sample of individuals with knee OA. Further study is needed to determine whether there
Post a Rapid Response to
are subgroups of individuals who might achieve an added benefit with this approach.
this article at:
ptjournal.apta.org

452 f Physical Therapy Volume 91 Number 4 April 2011

• Audio Abstracts Podcast been shown to be correlated with deficient knees might be modified to This article was published ahead of reduced functional ability. Agility and Perturbation Training Techniques in Exercise Therapy E xercise therapy is well estab. In addition. challenges to to activities that challenge the knee niques using walking-based rather balance. Adding the perturbation techniques. muscle weakness.11 The pro- programs may be effective in improv. it appears with knee instability. quick stops. negotiating obstacles) that to potentially destabilizing loads dur. knee instability. atic reviews indicate that the effects functional gains with exercise ther. The if individuals with knee OA were these types of training techniques to agility exercises focused on expos- better prepared to deal with these standard rehabilitation programs was ing the individual to quick stops and challenges to motor function. ptjournal. the use of lower-extremity strength. challenges of motor function (eg.10 challenges in balance. negotiating obstacles) that may be of exercise on pain and function are apy. these tech- Based on these findings.1– 4 Although these stops and starts. turns. 2011. and reduced training techniques involve quick based exercise therapy program for aerobic capacity. Although exercise therapy has been ited joint motion. Perturba. and The idea is that exposing individuals included performing agility tech- changes in direction. been shown to reduce complaints of complain of knee instability. turns. exercise programs should include encountered during daily functional moderate for people with knee OA. perturbation techniques in conjunc- as lower-extremity joint motion def.apta. Therefore. in people with ACL- sure to other challenges of motor challenge balance and knee stability. function in people with knee OA have complaints of knee instability. than running-based activities and may be encountered during daily ing therapy may help them learn to beginning in double-limb support functional activities.11 The people with knee OA may complain impairment-based exercise programs of knee instability. tion training techniques. The modifications function (eg. knees.org prevalence of self-reported instabil. individuals were exposed to such return to high-level physical activity The perturbation training involved challenges in motor function in con. with mod- provide the individual with expo. We reported on successful implementation of It also is well recognized that some Given the limited effects with this program in a case report.org. focusing on impair. than an impairment-based standard the use of rollerboards and tilt- junction with traditional impairment. and changes in direction. starts.7 This evidence quick stops. at is evidence that complaints of knee OA.7–9 Patients usually on improving physical function in describe this knee instability as “giv. may indicate that in order to enhance direction. quick turns and changes in improvement in overall physical ducing dynamic knee instability (or direction. challenges to balance. knee OA. Overall physical deal with these loads when encoun. Agility tion with a standard impairment- icits. people with knee OA.5.1– 4 quadriceps muscle weakness.7–9 There be beneficial for people with knee print on February 17.apta. cutting and turning. we reasoned that a ity among people with knee OA has similar approach that was found to • The Bottom Line Podcast ranged from 11% to 44% and has be effective for people with ACL. The with knee OA have complaints of ptjournal. niques provide exposure to other with knee osteoarthritis (OA). even if they do not programs. instability contribute to reduced viduals with knee OA who did not April 2011 Volume 91 Number 4 Physical Therapy f 453 . knee instability in some people with Exercise therapy programs for knee anterior cruciate ligament (ACL) Recently. even for those indi. injury. deficient knees. and negotiating obstacles function might be accomplished if improving dynamic stability) upon encountered in the environment. they do not tion training incorporates the use of activities that were used. recent system. This found to be more effective in re. rather than single-limb support for function might be further improved tered in regular daily activity. rollerboards and wobble boards to ifications. that included the use of agility and ments associated with knee OA such ening as well as agility and perturba. lished as an important part of above and beyond what is explained agility and perturbation training tech- clinical management for people by impairments such as knee pain. gram consisted of many of the same ing these impairments. program in people with ACL-deficient boards to expose the individual to based exercise therapy programs. niques may be beneficial to people that there is room for improvement with knee OA by enhancing func- in designing OA exercise therapy Therapeutic exercise approaches have tional ability. coupled with the fact that Available With ing way” or “buckling” of the knee a significant proportion of people This Article at during activities of daily living.10 These approaches included mental approach to exercise therapy based in design. we developed a supple- OA traditionally have been impairment. and changes in shown to be helpful. and lim.6 activities that will address problems activities.

and the University of Pitts- function and knee stability in patients with knee osteoarthritis. Sports Medicine. Because previous ity secondary to her knee OA and niques to a therapeutic exercise pro. has helped improve the outcome for young athletes who sustained vania. com- This study tested whether adding agility and perturbation training to a munity flyers. improved exercise therapy program without an additive effect in improving func- her Western Ontario and McMaster agility and perturbation training for tion. current study to determine whether that there would be a greater pro. tibiofemoral joint. participants in both groups appeared to improve in self-reported physical All participants provided written function. treatment group. volunteers who were recruited from What new information does this study offer? the greater Pittsburgh metropolitan area through physician offices. posttraining score⫽13). Method score⫽21) and her WOMAC physical ity and perturbation training tech. We hypothe- by 60% (pretraining score⫽5/10 training to an exercise therapy pro. was to formally test the effectiveness also examined the effect of the inter- gram. studies have shown that exercise can had to stop playing golf and tennis. help reduce pain5. ized clinical trial. Setting and Participants The Bottom Line The study was conducted in the Outpatient Department of Physical Therapy at the University of Pitts- What do we already know about this topic? burgh Medical Center. treating therapists instability. it is likely it would also have an Universities Osteoarthritis Index people with knee OA in a random.6 and that pain has She also had difficulty climbing stairs a strong relationship with physical without her knee feeling unstable. The purpose of the current study function in people with knee OA. Participant recruitment began an anterior cruciate ligament injury. We hypothesized training score⫽27. portion of participants reporting consent prior to enrollment in the ing function and self-reported knee improved knee stability in the group study. More work is needed to determine which patients rence radiographic changes in the may benefit from this approach. All participants gave written informed there is an additive effect on improv.Agility and Perturbation Training Techniques in Exercise Therapy patient in the case report had knee stability and reducing pain by adding who received the agility and pertur- pain and complaints of knee instabil. We also hypothesized dures) randomized controlled trial. Participants were person with knee osteoarthritis. bation training. she had reduced her knee pain of adding agility and perturbation ventions on knee pain. ticipate in the study. additive effect in reducing knee pain. The study was a prospective. Center for Agility and perturbation training. not participate in any testing proce- case prompted us to conduct the cise program. when added to standard exercise pro. sized that if the agility and perturba- and posttraining score⫽2/10 on 0 –10 gram in comparison with the same tion training program were to have numeric knee pain scale). newspaper advertise- standard exercise program could provide additional improvements in ments. dard exercise therapy program would blinded (testers were blinded to and no longer had complaints of knee have greater improvements in phys. adding agility and perturbation training ical criteria for knee OA12 and had techniques to your therapeutic exercise program may or may not increase grade II or greater Kellgren and Law- the benefits of therapy. Pittsburgh. If you’re a patient. what might these findings mean Individuals were included in the for you? study if they met the 1986 American College of Rheumatology (ACR) clin- If you have knee osteoarthritis. Pennsyl- grams. She was able to return to ical function compared with those were not blinded to group but did golf and tennis. there did not appear to be an additive effect of the agility and approval from their physician to par- perturbation training program. agility and perturbation training tech. and the final data collection procedure was performed approach also might improve the effects of therapeutic exercise in a in December 2008. we Following 12 sessions of the pro. (WOMAC) total score by 22% (pre. posttraining that participants who received agil.13 The 1986 ACR criteria for diagnosis of knee OA 454 f Physical Therapy Volume 91 Number 4 April 2011 . Although burgh Arthritis Institute Registry. One case report suggests that this in October 2004. The success of this who received only the standard exer. single- score⫽17. gram for people with knee OA. Design Overview function score by 24% (pretraining niques in conjunction with a stan.

randomization was strati.12 Individu. muscle stretching (quadriceps femo. The maintain balance and control over the supervised sessions. Following faces. individuals were excluded if they had undergone total Participants assigned to the agility To account for any potential effect knee arthroplasty. the exercises. were unable to ambu. braiding (lateral stepping com. block randomization was back crossover steps during back. Details of both the following: age 50 years or older. utes’ duration. addition of agility and perturbation pants in the standard exercise group extremity function. pists were given one-on-one instruc- gins of the joint. hamstring. backward and forward) during walk. ing. tiltboards. To ensure standardization of the inter- ting while weight bearing). modifications. performed an arm-bike exercise ing techniques included side step. the a drill requiring multiple changes period. The participants attempted to the program they received during sealed envelope in sequence. long-sitting knee flexion of the participants’ treatment they required use of an assistive and extension range of motion.4 therapists every 6 months. diagonally participants in the home exercise sequentially numbered. by the principal investigator (G. This amount of and Interventions bined with forward and backward time approximated the time it took To ensure a balanced assignment of crossover steps).2. activities so that they would be inde- lopes containing participant inter. To eliminate any potential biases ward ambulation. and calf muscle stretch. Agility and Perturbation Training Techniques in Exercise Therapy include knee pain and at least 3 of any of the follow-up testing or inter.to 8-week assignment in the block sequence. exhibited uncon. agility and perturbation training used. lems. reported a treadmill walking. tender. grams for knee OA. crepitus with active Participants were randomly assigned motion of the knee (eg. sions. ing) and strengthening (quad sets. The therapists were provided with a graphic changes were included pro. and rollerboards to the study. ing the perturbations. partici- logical disorders that affected lower. The content of the home baseline testing. als with patellofemoral joint radio. and progressing programs per the previous year. Participants were to com- that could occur from an investigator (forward and backward walking to plete 12 supervised sessions of their determining the next intervention and from designated markers). and standing conducted periodic. The principal investigator late a distance of 30. bony enlargement. had a history of same standard exercise program as therapy between groups may have cardiovascular disease. of these exercise programs are pro- morning stiffness of less than 30 min. gram that included lower-extremity tion in the intervention procedures and no palpable warmth. The trial coordinator safely participate in the training pro. booklet that included descriptions vided that they had tibiofemoral supine straight leg raises. or had neuro. A set of right and left lateral steps.). with some trial coordinator did not take part in the exercised lower extremity dur. vention procedures. of the supervised therapy period of by the study statistician. The perturbation techniques pendent in this program by the end vention assignment were created incorporated the use of foam sur. Those assigned to the standard exer. To hip extensions. to 1 of 2 exercise intervention groups. the standard exercise group with the had on treatment outcome. individuals were excluded if heel raises). cise group received an exercise pro. prone and pictures of the exercises and radiographic changes as well. This program was consistent protocol. seated isometric instructions in the progression of ensure that study participants could knee extensions. and respective program in a 6. All lower-extremity receiving their assigned protocols history of 2 or more falls within the exercises were performed bilater.F. depending upon their sched- intervention assignments occurred in in direction in which the therapist uling constraints. prompt the individual to change During the supervised exercise ses- fied by the presence of unilateral direction (forward and backward.K. training techniques. 10 to 15 minutes. when squat. standing hamstring curls. and perturbation group received the that differences in contact time in trolled hypertension. shuttle walking activities. therapists began instructing versus bilateral knee OA. single-leg leg presses. random block sizes of 2 and 4. activity of the upper extremities for Randomization ping. In addition. ris. ally. random reviews grams. and records to ensure that they were device for ambulation. vided in the Appendix. front crossover for participants in the agility and participants to the 2 intervention steps during forward ambulation. or reported severe visual prob. In provided hand signals at random to addition. the trial coordinator expose the individual’s lower limbs exercise program for the standard assigned an individual to a group and body to potentially destabilizing exercise group was similar to that of based on the instructions in the next forces. The agility train. perturbation group to complete the groups.1. the intervention programs with the period. all thera- ness on palpation of the bony mar. sealed enve. vention implementation. Because they would April 2011 Volume 91 Number 4 Physical Therapy f 455 .5 m without with current published treatment performed a face-to-face review of an assistive device or need of a rest recommendations for exercise pro.

informed of the specific group to press. No outcomes were tested at scores separately from the total leg standing balance exercise on this 4-month visit. ments. and Get Up and participation through the 6-month mented through the 6-month follow. There would be no penalty OA. each week. The Participants in the agility and pertur. and 17 items related to walking on therapist command. The reliability and validity of would be safety concerns in per. In addition. Go Test (GUAG) score. wall squat exercises were participants were encouraged to con. Participants rated the severity of OH 44310-2575. itoring was not performed. much they actually did their exercise function for individuals with knee formed all agility exercises with the program. 1245 Home Ave. A number of steps were taken to The physical performance measure- lated to the wall squat. hip extensions. 6. and at 2. The same standard home program exer. exercise could not be directly trans. Self-report measure- mum resistance used in the leg press ments were taken at all time points. they per. Akron. self- exercises at least 2 times per week.1) could not provide them with this This visit was to ensure that partici. them that they would be paid for WOMAC total score was the primary bation group performed all of the returning the diary regardless of outcome measure for the study. They were encour. and hamstring curls at home. they were not Band* (Gold color) and instructed monetary reward for returning their involved in any of the study testing how to attach it to a chair in order to exercise diaries at the follow-up visit. and physical cise group. not be blinded from the participants’ provided with heavy-resistance Thera. encourage adherence to the home ments were taken at baseline and pants had been lifting about half of exercise program. They visit with the study coordinator for a physical function. Although a 1-repetition maxi. Finally. difficulties with the home program. their body weight during the leg provided with a binder containing Testers were blinded from knowl- press exercise. procedures and remained blinded perform an isometric knee extension In order to prevent any bias from from test results. study. we instructed Primary outcome measure. gram. global rating of We monitored their home program exercise programs14 and was imple. scribed exercise program properly WOMAC scores have been estab- forming these activities without and to assist them in troubleshooting lished.Agility and Perturbation Training Techniques in Exercise Therapy not have a leg press machine in their follow-up period. stiffness components in the WOMAC ticipants in both groups also were phone contacts to remind partici. Participants were at the 2. but further mon. The participants were not resistance similar to that of the leg lists were provided to the partici. The WOMAC comprises 5 items exception of the activity involving for incomplete forms. For the returned at each follow-up visit Although the treating therapists could isometric knee extensions. change (GRC) score. As an alternative. that we only wanted to monitor how sure of pain. most partici. Outcome Measures homes. they were through the 6-month follow-up visit. up period. They with cuff weights to perform the amount of exercise they performed only knew that they were randomly straight leg raises. These diaries were to be assigned to 1 of 2 exercise approaches. and 12 months following cise. the trial scores to ensure that the pain and level surfaces and carpeting. coordinator conducted monthly tele. knee instability on a 0 to 5 numeric 456 f Physical Therapy Volume 91 Number 4 April 2011 . Participants were provided a small group assignments.15–17 We also analyzed the supervision. Participants also related to pain.and 6-month time points. total score did not mask the potential encouraged to continue a walking pants to perform their exercises effects on physical function. whether the form was complete and WOMAC is a disease-specific mea- cises as those in the standard exer. they any problems with the home pro. Exercise diary check. Secondary outcome measures were: aged to perform all prescribed home This multistep approach has been self-reported knee instability. of the WOMAC was used in this equipment for home use and there pants were performing their pre. After this period. 0 to 4 Likert version (version 3. Each item is also did not perform tiltboard and review of the home exercise pro. 2 items related to multiple changes in direction during came in to the clinic for a face-to-face stiffness. WOMAC physical function subscale were instructed to perform a single. randomization. stiffness. participants forging their diaries to receive the reward. * The Hygenic Corp. Participants were provided pants so that they could record the which they were assigned. tinue the program. exercise. home program. scored on a 5-point Likert scale. program of at least 30 minutes per and complete their exercise diaries day at least 3 days a week for the and to assist in troubleshooting any Secondary outcome measures. shown to improve adherence to reported knee pain. Par. Testing was performed at baseline substituted for the leg press exer. so we were confident descriptions and pictures of their edge of participants’ group assign- that the wall squats were providing home exercises. The rollerboard activities because we gram 4 months post-randomization.

we my activity slightly”. serious adverse events.18 The test-retest reliability GUAG. associated with the interventions or rate the degree to which their knee comorbidities.” 8 corresponds to the phrase “about the same.98) test-retest into an effect size of approximately come Survey—Activities of Daily Liv. 2⫽“the symp.” took to complete the task.26). and interrater (ICC⫽. Questionnaire physical activity sub. we com. Because few par.44. therefore. race. based on preliminary your level of daily activity?” The rat.22 The minimum detectable sidered a moderate treatment effect of this self-report rating of knee change was 1. and chant view for the patellofemoral tures. great deal better.1]⫽. The GUAG was used as a performance. unobstructed mean between groups with a com- 1⫽“the symptom affects my activity distance of 15. with knee OA undergoing exercise have the symptom”. Our analysis indi- ticipants rated their instability as 0 or trolled in the final analysis. observations in our clinic of people ings were as follows: 5⫽“I do not based measure of function.95) hypothesis. graphic characteristics. used to measure the length of time it on the WOMAC. reliability in our laboratory for the 0. we needed to enroll sion (measured using the Center for 240 participants (120 per group) to Participants rated the worst knee Epidemiological Studies—Depression account for this loss. A stopwatch was mon standard deviation of 18 points severely”. exercise between treatment groups 1 corresponds to the phrase “a very (Tabs. modified for the knee25. testing procedures that required refer- condition has changed from the number of dropouts. This estimate translates bility was taken from the Knee Out.13 knee alignment (using the adverse events (injuries or symptoms 15-item scale in which individuals long cassette radiographic view). Summary statistics are presented as with 0 representing “no pain” and 10 scale.19. We anticipated that approxi- ⱕ3). mately 30% of the participants might if the knee instability rating did not medications. pain they experienced in the 24 Scale24). participants were therapy and those who had received symptom. Assessment of additional agility and perturbation ing knee OA.22 between groups. We have significance and a 2-sided alternative This self-report rating of knee insta. physical activity (measured using the ables and as percentages for discrete inable. cated a sample size of 168 partici- 1.2 seconds. body mass index. we collapsed the ratings into a pared potential differences between pants (84 per group) would yield dichotomous variable. Parametric and nonpara- been found to be reliable and valid Elderly27) (Tab. radio. and adherence to home present.” Numeric rating scales have Physical Activity Scale for the variables.05 level of vents me from all daily activities. with a . and adherence ral to a physician for evaluation or time treatment was initiated to the to the home exercise program were intervention). includ. fear of physical activity (mea- hours prior to testing using an sured using the Fear-Avoidance Beliefs Data Analysis 11-point numeric pain rating scale. metric analysis-of-variance methods for measuring clinical pain.” and 15 April 2011 Volume 91 Number 4 Physical Therapy f 457 . point. they stood up and walked as fast as assumed an 8-point difference in the tom affects my activity moderately”. On the command “go. To per. 1). a rating of assessed (Tabs. 2 and 3. demonstrated intrarater (ICC⫽. 4⫽“I have the form this test. anxiety (measured using either fail initial screening or drop affect their activity (ie.2 m. Fig.20 graphic severity of knee OA (using and chi-square tests were used to com- the standing fixed knee flexion view pare the baseline clinical and demo- A GRC score was obtained from each for the tibiofemoral joint and Mer. They were classified as “stable” years with a diagnosis of knee OA. covariates that might need to be con. and we believed it instability on 50 individuals with a would be reasonable if adding the variety of knee pathologies. The GRC is a joint). great deal worse. In addition. or shifting of the knee affect for the study. Participants groups on a number of variables that approximately 80% power to detect were classified as “unstable” if their could influence treatment outcome. height. buck. This effect size would be con- ing Scale. depres. was estimated using Potential Covariates training to existing exercise therapy an intraclass correlation coefficient In order to account for any potential programs was to be worth the extra (ICC [2. but it does not affect daily seated on a standard-height chair agility and perturbation training as activity”.21 For example. weight. out. and 0⫽“the symptom pre. 1). 6-. sex. time and effort. 1. 2.” part of the exercise program. Agility and Perturbation Training Techniques in Exercise Therapy scale in response to the query: “To corresponds to the phrase “a very Sample Size what degree does giving way. treatment fea- participant at the 2-. 3⫽“the symptom affects with armrests. WOMAC scores at the 1-year end- enough to affect their activity (ie. and means (⫾SD) for continuous vari- representing “the worst pain imag. and 3). ⱖ4). rates of side effects and serious 12-month time points.” To estimate the sample size required ling. a difference of 8 in the mean knee instability rating was severe These variables included age. the Beck Anxiety Index23). possible along a level.72).

b n (%) unstable 28 (30.4) 60 (65.2) .83 Knee instability rating.7 (5.0) 26 (28.2 (6.5) .0) Medications at baseline.18 Black 10 7 .1) .6) . X (SD) 141.1 (16.7) . n (%) 35 (38.7 (2.85 Corticosteroid 2 (2.1) 30. X (SD) 19.3 (8.4) 5 (5.4 (2. n (%) .16 Native American 2 1 .7) 28.1 (15.3 (6. X (SD) 9.2 (86.0) .5) 1–2 11 (12. X (SD) 4.3 (18.5) 15 (16.6) 20 (22.7) .6 (2. n White 78 83 .0 (5.5 (7.6 (8.85 Glucosamine 14 (15.1) 12 (13.2) 162.6) .36 ⬍1 5 (5.2) 23 (25.86 a WOMAC⫽Western Ontario and McMaster Universities Osteoarthritis Index.6 (8. n (%) Analgesic 13 (14.59 Beck Anxiety Index score.6 (2.8) 32 (35.94 Center for Epidemiological 7.95 WOMAC physical function score.08 Numeric knee pain rating.15 physical activity score.6) .31 Sex (female).20 Get Up and Go Test score (s). Baseline Characteristics by Treatment Group Agility and Standard Exercise Group Perturbation Group Characteristic (nⴝ92) (nⴝ91) P Age (y).6 (67.6 (20.81 Body mass index (m/kg2).0) 39 (42.1) 5–10 25 (27.83 Race.1 (7.9) .51 Years with arthritis.9) 19. X (SD) 168.2) 7.1) .8 (6.64 Height (cm).1) 167.1) 3–5 21 (22.5) .9) .3) 4.46 Hispanic 0 0 . X (SD) 4.4) 36 (40. 458 f Physical Therapy Volume 91 Number 4 April 2011 . X (SD) 85.0) 86. X (SD) Physical Activity Scale for the Elderly.9) .3) 9. X (SD) 30.8) .3) . ⱖ4⫽stable.73 Studies–Depression Scale.4) 63. b ⱕ3⫽unstable.72 Nonsteroidal 23 (25.86 Injection 17 (18.5) .0 (10.4) 4.7) 11.2) 1 (1.0) .9 (11.0) 11 (12.8) .2) 13 (14.39 a WOMAC total score. X (SD) 64. n (%) 62 (67.57 Weight (kg). X (SD) Fear-Avoidance Beliefs Questionnaire 10.99 Asian 2 0 .3) ⬎10 30 (32.5 (12.Agility and Perturbation Training Techniques in Exercise Therapy Table 1.1) .57 Prior history of knee injury.3) .7 (5. X (SD) 28.0) 11 (12.57 COX-2 inhibitor 12 (13.

3) Grade III 24 (26.3) 22 (24.1) 28 (30.3) .0) .5) Grade II 28 (30.7) Grade I 26 (28.6) Grade IV 9 (9.8) 8 (8.5) Unilateral involvement 10 (10.5) 44 (48.7) Grade I 18 (19.8) Grade II 29 (31.6) 20 (22.8) 11 (12.19 Grade 0 3 (3. varus (°).58 Grade 0 9 (9.4) 21 (23.7) Grade IV 4 (4.1) 178 (3.1) 29 (31.3) Grade I 10 (10.2) 5 (5.7) Grade I 14 (15.7) Grade III 19 (20.7) Grade II 19 (20.13 Grade 0 3 (3.3) Grade III 24 (26.8) . X (SD) 177 (4.2) 18 (19.45 Most-affected knee alignment.9) 13 (14.3) 6 (6.7) 27 (29.7) 34 (37.1) Least-affected knee patellofemoral compartment .5) Most-affected knee patellofemoral compartment .6) Grade I 19 (20. varus (°).8) 8 (8.9) Grade IV 13 (14.4) 5 (5.3) 3 (3.1) 34 (37.2) Grade II 19 (20.7) 29 (31. Baseline Characteristics by Treatment Group: Radiographic Severity and Comorbiditiesa Agility and Standard Exercise Group Perturbation Group Characteristic (nⴝ92) (nⴝ91) P Radiographic severity (Kellgren and Lawrence grade) Least-affected knee medial compartment .5) 23 (25.6) Grade IV 9 (9.3) Grade IV 22 (23.9) 18 (19.92 (Continued) April 2011 Volume 91 Number 4 Physical Therapy f 459 .1) 9 (9.2) 31 (33.5) Grade II 37 (40.7) 27 (29.71 Grade 0 10 (10.7) 12 (13. Agility and Perturbation Training Techniques in Exercise Therapy Table 2.8) 17 (18.3) Grade II 34 (36.7) Grade I 17 (18.5) Most-affected knee lateral compartment .5) 17 (18.73 Grade 0 24 (26.9) Grade III 32 (34.3) Grade III 21 (22.8) 5 (5.7) 23 (25.8) Least-affected knee lateral compartment .5) 176 (5. X (SD) 176 (5.9) Most-affected knee medial compartment .3) Grade IV 12 (13.3) 7 (7.9) 5 (5.23 Grade 0 15 (16.1) Grade III 29 (31.49 Least-affected knee alignment.

Home Exercise Adherence by Group Agility and Standard Exercise Group Perturbation Group n (%) n (%) P a Session adherence .7) 7 (7. we adjusted ment groups. as it was significantly different logistic for discrete measures) were results.36 Memory problems 8 (8.7) . totaling 32 sessions. approach to missing data.1) 1 (1.4) 7 (7. The expectation was 2 sessions per week over a 16-week follow-up period. Furthermore.12 Back pain 47 (51.13 Heart disease 8 (8. We adjusted the (ITT) principle.1) 0 (0.1) 2 (2.7) .0) . we conducted sensitivity between groups at baseline.34 Depression 20 (21.4) .63 Adherent (⬎80%) 48 (52) 43 (47) Partially adherent (50%–80%) 15 (16) 19 (21) Not adherent (⬍50%) 26 (28) 22 (24) No adherence data 3 (4) 7 (8) Total 92 (100) 91 (100) a Based on the percentage of expected home exercise sessions to complete during the follow-up period.99 Diabetes 4 (4. To determine whether this models for the incidence of can- (linear for continuous measures and imputation approach affected study cer. ized estimating equation longitud. Continued Agility and Standard Exercise Group Perturbation Group Characteristic (nⴝ92) (nⴝ91) P Comorbidities High blood pressure 38 (41.2) . At the 12-month assess. Activity Scale for the Elderly score on the outcome variables at the Consistent findings indicated that because although not statistically 12-month follow-up assessment.7) 7 (7. unless otherwise indicated. adjusted the models for the Physical was an independent treatment effect cated using multiple imputation. Regression models approach.3) .80 Previous hip fracture 0 (0.41 Cancer 19 (20.2) 0 (0.99 Liver disease 2 (2.80 Stomach ulcer 7 (7. Table 3.7) . ences appeared to approach signifi- teristics that differed between treat.7) 12 (13.7) 8 (8. data for participants with the association of treatment on the compartment and patellofemoral 460 f Physical Therapy Volume 91 Number 4 April 2011 . the group mean differ- adjusting for those baseline charac.02 Congestive heart failure 1 (1.98 Stroke 1 (1.50 Lung disease 7 (7.8) .62 a Values represent number of participants (%). cance.2) .7) .1) .25 Kidney disease 1 (1. We also used to determine whether there analyses.1) .0) . The analyses were repli.1) 41 (45.0) 2 (2. the results were not affected by this significant. General. All outcome analyses were conducted missing scores were imputed using outcomes over the course of the according to the intention-to-treat the last observation carried forward follow-up period.6) 4 (4.3) 48 (52.6) 7 (7.8) .Agility and Perturbation Training Techniques in Exercise Therapy Table 2. inal models were used to model for the most-affected knee medial ment.

1 and 2. PT⫽physical therapy. so the results were enrolled in the study. eight participants were not random. there did appear to be contained in this publication are pants. 183 participants were April 2011 Volume 91 Number 4 Physical Therapy f 461 . UKA⫽unicompartmental knee arthroplasty.28 No adjustment for Results cancer in the list of comorbidities. Forty. nificantly more participants with should be interpreted accordingly. by group. 1). ing from participation in the study National Institute of Arthritis and Therefore. are shown in Tables some difference between the groups those of the grantees and do not nec. 2). multiple tests was implemented for Two hundred thirty-one participants The standard exercise group had sig- secondary endpoints. because. compartment radiographic severity Musculoskeletal and Skin Diseases randomized to a treatment group. The views The characteristics of these partici- significant. CONSORT diagram of study enrollment and participation. Agility and Perturbation Training Techniques in Exercise Therapy Enrolled Enrollment (n=231) Excluded (n=48) Did not meet inclusion criteria (n=34) Refused to participate (n=10) No PT referral from MD (n=4) Randomized (n=183) Allocation Standard Exercise Group Agility and Perturbation Group (n=92) (n=91) Lost to follow-up (n=1) Lost to follow-up (n=2) TKA (n=1) THA (n=1) Illness (n=1) Refused futher participation (n=1) Missed 2-month testing visit (n=5) Missed 2-month testing visit (n=12) Follow-up 2-Month Standard Exercise Group Agility and Perturbation Group (n=84) (n=75) Lost to follow-up (n=3) Lost to follow-up (n=5) UKA (n=2) UKA (n=1) Illness (n=1) THA (n=1) Refused further participation (n=2) Illness (n=2) Missed 6-month testing visit (n=3) Missed 6-month testing visit (n=3) Follow-up 6-Month Standard Exercise Group Agility and Perturbation Group (n=78) (n=75) Lost to follow-up (n=3) UKA (n=3) TKA (n=2) Lost to follow-up (n=1) Illness (n=3) Missed 12-month testing visit (n=1) Death (n=1) Refused further participation (n=3) 12-Month Analysis Completed study (n=69) Completed study (n=76) All 92 in ITT analysis All 91 in ITT analysis Figure 1. variables or potential covariates. as they did and perturbation group (Tab. ITT⫽intention to treat. a history of cancer than the agility ized to a treatment group. although not statistically (grant 1-R01-AR048760). There were no differences on these variables and previous essarily reflect those of the funding between groups for any demographic research indicated that radiographic agency. severity might affect the outcome of with the exception of history of exercise therapy. TKA⫽total knee arthroplasty. Role of the Funding Source not meet the study inclusion criteria There were no adverse events result- This study was supported by the during screening procedures (Fig. THA⫽total hip arthroplasty.

(3.5–4. (Treatment (Treatment Measure n n n n n n n n Unadjusted Adjustedb Effect) ⴛ Time) WOMAC total 22. (2.7 11.35 . group.9 (29/91) .1 3.8). n⫽84 n⫽75 n⫽78 n⫽75 n⫽66 n⫽76 n⫽92 n⫽91 Get Up and Go 9.3 (22/75) 21.55 a 95% CI⫽95% confidence interval.5–22.7–6.4–4.9–5. (95% CI).52 . (4. (95% CI).7).7–19.1 19.8 4.2 8.9 20.8 16.4–15. these differences dard exercise group in WOMAC total more improvement than the stan- turbation group exhibiting slightly baseline.4 5.4–5.0).3).6–27.1–4.6). (16.04 function score (12. the results of the ITT analysis.1 5. (10.4–9.6–6.1–18.1–18.74 .6 (21/76) 26.7–9.2).2 12.5–26.8–4. (3.5–9.5).2).1 (24/92) 31. scores over the follow-up periods to WOMAC total and physical function We have provided a plot of the GRC scores in the early follow-up and physical function scores and disappeared by the 1-year endpoint.1 . (4. (4.6–6. both groups exhibited some modest who had knee instability within the in the proportion of participants or GUAG scores within the groups. we scores for each group to percentages been reported to be clinically mean- WOMAC scores from baseline has nificant.67 .3). (4. b Adjusted for incidence of cancer.4–9.3).8 9. (8.4). (95% CI).9 13. (3. (19. (95% CI). (95% CI).9–11.4 .9 9.31 .0–4. (8. (10. (3. n⫽80 n⫽74 n⫽69 n⫽70 n⫽92 n⫽91 Knee instabilityc 30.5).7). slightly more than the standard exer- and perturbation group improved Although we did find a significant interaction indicating that the agility between the intervention groups. A 17% to 22% change in Our results indicate that although ingful. c Agility and Perturbation Training Techniques in Exercise Therapy Reported as percentage of participants with knee instability (ⱕ3 on the knee instability rating scale). (2.7).9 15.13 Number 4 (3.03 change (4. (95% CI). (15.3 4. (3. (8.0–4.0 (25/84) 25. (95% CI).4 (19/78) 29.8–10.9–5.2 . n⫽84 n⫽75 n⫽78 n⫽75 n⫽66 n⫽76 n⫽92 n⫽91 Knee pain 4.6).3).9).1). The significant treatment ⫻ come measures at the 1-year end- treatment groups for any of the out- there were no differences between gitudinal ITT analysis indicated that Table 4 provides the mean outcome for each outcome variable.3 23.0 (19/75) 24.6).1 . Physical Activity Scale for the Elderly score.8–4.2).4 5.4).6 3.7–27. n⫽84 n⫽75 n⫽78 n⫽75 n⫽66 n⫽76 n⫽92 n⫽91 Volume 91 WOMAC physical 15. with the agility and per- improvement in both groups from Table 4 indicate there was some time interaction results shown in point.7–17.1–23. (10.5–4.7–25. Discussion post hoc testing of the simple April 2011 of change in WOMAC scores.7–6.5 5. (15.05 score (18. groups.0 19.0 23. 462 Outcome Measures by Treatment Group at Each Time Point With Intention-to-Treat (ITT) Analysisa f 2-Month Follow-up 6-Month Follow-up 12-Month Follow-up 12-Month ITT Sample Treatment P Values Standard Agility and Standard Agility and Standard Agility and Standard Agility and Exercise Perturbation Exercise Perturbation Exercise Perturbation Exercise Perturbation Group Group Group Group Group Group Group Group 12-Month Follow-up Longitudinal Adjustedb Adjustedb Mean Mean Mean Mean Mean Mean Mean Mean Physical Therapy (95% CI).3).57 . improvements in outcome measures.3).62 .1 13.6 19.7).0).6). (8. (3.6).1–24.56 .3–15. WOMAC⫽Western Ontario and McMaster Universities Osteoarthritis Index.6–6.89 .4).9 23.4 3.4–22.40 .45 . (16.3).0). (13.90 . period.1).96 . (13. Table 4.1–16.32 . (19. 2).5 3.56 .5). (19.6 13.75 Test score (8. most-affected knee medial and patellofemoral compartments radiographic severity. (11. (9.5 5. (3. The lon- and the treatment ⫻ time interaction point.4).1 8. (16. there were no significant differences There also was no significant change There were no changes in knee pain illustrate this interaction (Fig.0).8–12.6 13.29 By converting our change between-group effects were not sig- cise group at early time points.4).2).3 4. scores for each group at each time reported by participants in either .3).2 (14/66) 27.5 .39 .3). However.5 4. n⫽84 n⫽75 n⫽78 n⫽75 n⫽66 n⫽76 n⫽92 n⫽91 Global rating of 5. (4.0 3.

Asterisk indicates both groups’ scores were significantly different from their respective baseline scores (P⬍. Agility and Perturbation Training Techniques in Exercise Therapy Figure 2.01). Plot of Western Ontario and McMaster Universities Osteoarthritis Index total (A) and physical function (B) scores at each time point. April 2011 Volume 91 Number 4 Physical Therapy f 463 .

but perhaps the gram in our sample of participants had anticipated a 30% initial screen.5 seconds) in both groups. At the 1-year endpoint. time frame. majority of them did not. role limitations (physical). missing data. The GRC scores also faster) following treatment. we tive devices for ambulation.30 Dira. and tocol data using a Bonferroni correc. Therefore. agility and perturbation group⫽ ence was on the order of 0. The results do not Our initial sample size estimate indi.3 to 1. We did not meet the target we had excluded participants who study that examined the effective. balance. clude that the groups were not dif- was within or exceeded this range tically better in the kinesthesia ferent on our outcome measures. (2 months: standard exercise group⫽ group. vey questionnaire (SF-36) physical direction. the actual differ- 24%. but it is younger. the relatively group⫽29%.7 points. difference in WOMAC scores. and as we and agility may occur. participants received 24 sessions of 1-year endpoint. They hypothesized that exposing partici- group⫽33%. quick changes in pair-wise comparisons on the per pro. so on. cient statistical power. injuries. agility and per. agil. the daily chal- to 4⫽moderately better) at all 3 ence. function. The kinesthesia group also ment problems during daily function. Likewise. Another potential limitation is that We are aware of only one other pants. ticipants engaged in higher-level and perturbation training techniques pants (84 per group) to have suffi. groups indicates that we probably centage of change from baseline in coglu et al did report that WOMAC did have sufficient power to con- WOMAC physical function scores physical function scores were statis. as we enrolled only 231 were already known to be at risk for ness of a similar training program participants due to limitations in falling and those who required assis- that included balance and agility funding. 2). We 6 months: standard exercise group⫽ an 8-week period. because we used an benefit more from the agility and per- tiveness of “kinesthetic and balance ITT analysis. as frequent as those encountered by pain (0. with knee OA. activity where challenges to balance to a standard exercise therapy pro. the per. although approach. 12 months: standard also reported greater improvements pants to challenging movement prob- exercise group⫽32%. we had group⫽17%. pants. sible that these individuals might Diracoglu et al30 compared the effec. in scores on the Medical Outcomes lems. There were colleagues’ study received twice as for agility may not be as intense or only slight improvements in knee many treatment sessions as our par. However. which is well over ity and perturbation group⫽32%. Participants in Diracoglu and lenges to balance and requirements follow-up time points.68 seconds and appeared to have worked in groups (Fig. ing failure or dropout rate. number of treatment sessions in a may not have as dramatic an effect Likewise. The data from 183 participants at the standard exercise group⫽21%. enrollment. It is pos- exercises in people with knee OA. we used a last score turbation training than from a standard 464 f Physical Therapy Volume 91 Number 4 April 2011 . In addition. It may be possible that adding agility and perturbation train- and we would not consider these if we had administered a greater ing techniques to exercise programs changes to be clinically meaningful. Outcome was performed a sensitivity analysis using 16%. in older individ- for each group (5⫽somewhat better sents a clinically meaningful differ. stability. agility and perturba. dealing with these types of move- changes from baseline in the WOMAC scales.10 However. In developing the agility and per- 34%. OA as had previously been observed knee instability from baseline to the for younger. athletic people. Similar to our study. athletic individuals. 12 months: standard exercise period. however.0) and in GUAG scores ticipants received and in a shorter younger. had complete data on 143 partici. would better prepare them for tion indicated that the within-group and vitality (energy or fatigue) sub. there were only slight shorter period of time. quick stops and starts.3 to 0. It follow-up periods. there was no small difference in means between tion group⫽28%). total and physical function scores had significantly faster times on the This logic may have been reasonable were significant (P⬍. both groups improved. so the analysis included range at most time points (2 months: cise in 60 female participants. (0.Agility and Perturbation Training Techniques in Exercise Therapy found that changes for each group training” plus strengthening exer. uals with knee OA. does not turbation training program. in our opinion. agility and perturbation seem clinically meaningful. athletic individuals with ACL exhibited moderate improvements unclear whether this finding repre. we might on knee stability and general func- within-group improvements in the have detected significant differences tion in older individuals with knee proportion of participants reporting between our intervention groups. carried forward approach to handle either were within or exceeded this cises versus only strengthening exer. may be true that some of our par- support an additive effect of agility cated that we needed 168 partici. agility and perturbation group⫽ measured at the end of the 8-week multiple imputation to verify our ITT 32%. including challenges to knee turbation group⫽29%). 6 months: standard exercise which. we indi- cated we would enroll 240 partici. Post hoc Study 36-Item Short-Form Health Sur. their respective interventions over the initial sample size estimate.01) for both 10-Meter Walking Test (0.

who are also at risk for falling may precision of decision making. ment response for both of our inter. Although both intervention groups have found differences between our 2 Jordan KM. et al. research design. Reisman AS. Osteoarthritis Cartilage. we might 2000. Further study is and perturbation training techniques gang provided institutional liaisons. Future cise for osteoarthritis of the hip or knee. In addition. 4 Zhang W. Wisniewski.2522/ptj. Dr Gil. Dr Oddis pro- may be that the addition of agility vided participants. Dahm KT. and Dr Irrgang provided writing. Phys Ther. our outcome measures may not have is likely to benefit from adding agil- been sufficiently sensitive in captur. exhibited some moderate improve. played a role in dampening any ben. 9 Schmitt LC. and it would function. whether baseline clinical measures 6 Jamtvedt G. whether there are patient character- istics that could be measured at base. 1 American College of Rheumatology Sub- we had only one performance-based vention groups. ing treatment. Exer- changes in knee pain from base. Dr Piva warranted to determine whether add. Dr Wisniewski. et al. Dr Piva. laxity. evidence-based approach to the manage- in a recently completed feasibility ment of knee osteoarthritis: report of a sures. 2003. Niu J.88:123–136.62:1145–1155. more ment of Osteoarthritis.147:534 –540. eficial effects of adding agility and could be used to identify individuals Physical therapy interventions for patients perturbation training to exercise pro. Christie A. 2008. line in both groups most likely also Cochrane Database Syst Rev. Bell M. balance. of joint instability in knee osteoarthritis: its prevalence and relationship to physical people with knee OA. Agility and Perturbation Training Techniques in Exercise Therapy exercise program. the WOMAC does not include but we are currently examining our DOI: 10. pro. we had responders provided project management.3: study also is needed to determine CD004286. Instability. ments in self-reported outcome mea. Phys Ther. Dr Gil and Dr Oddis provided data collec. Nuki G. measure of function. bation training techniques compared cal function measures. Arthritis Rheum. Dr Piva. Ann observed if there were not signifi. 2004. Arthritis Rheum. Dr Intern Med. 85:907–971. with osteoarthritis of the knee: an over- who may benefit from adding agility view of systematic reviews. function had been used. Dr Fitzgerald and Dr Irr- April 2011 Volume 91 Number 4 Physical Therapy f 465 . expert consen- sus guidelines. cant differences in knee pain follow. Clinical trial registration number: NCT00078624. and and perturbation training. Rudolph KS. 2003. not be surprising that differences 8 Felson DT. Piva SR. The fact that there were only slight lems with balance may be needed to 5 Fransen M. Ann Rheum tion training techniques in a stan. This effort This article was submitted May 26. part exercise program. function in those individuals. identify baseline predictors of treat- References ance and agility skills. ing aspects of physical function that niques to exercise programs and who might be more influenced by agility is not likely to benefit. are advantageous for some and not and Dr Gil provided clerical support.88:1506 –1516. we found Therapeutic Trials (ESCISIT). Moskowitz RW. Doherty M.20100188 items that directly incorporate data to determine whether we can higher-level activities requiring bal. and physical tion. Reports and function is well established in people with knee OA. 2007. received the same agility and pertur. EULAR Recommendations 2003. Dr Fitzgerald and nonresponders to treatment. and Dr Gil In both groups. For example. there was no additive benefit task force of the Standing Committee for study including participants with from including agility and perturba. an intervention groups. McConnell S.51:941–946. variety of performance-based physi. observe an additive effect. This study was supported by the National Institute of Arthritis and Musculoskeletal and Another potential problem is that line that would help determine who Skin Diseases (grant 1-R01-AR048760). Dr Fitzgerald. with knee OA who also have prob. If other types Recommendations for the medical man- agement of osteoarthritis of the hip and of performance-based measures of Conclusion knee: 2000 update. provided consultation (including review of techniques for people with knee OA for others. Dis. dard exercise program for our partic. For exam. 2008. et al. grams to improve physical function. Fitzgerald GK. 3 Ottawa Panel Evidence-Based Clinical sure of balance) and self-selected gait ipants with knee OA. buckling: prevalence. In order to improve the manuscript before submission). 2005. et al. Phys Ther. or perhaps harmful.43:1905–1915. strong relationship between pain to exercise therapy programs for 7 Fitzgerald GK.31 interventions to subgroups of people II: OARSI evidence-based. McClennan C. was beyond the scope of our study. It provided fund procurement. Dr Piva. and in physical function would not be and Dr Irrgang provided concept/idea/ associated limitations in function. Arden NK. Dr Gil and Dr Wisniewski provided data function in patients with medial knee osteo- arthritis. with those who received a standard OARSI recommendations for the manage- prioception) or application of the ment of hip and knee osteoarthritis. analysis. Dr Fitzgerald. Knee Dr Fitzgerald. Irrgang JJ. committee on Osteoarthritis Guidelines. was accepted December 28. The use of Practice Guidelines for Therapeutic Exer- speed improved in participants who cises and Manual Therapy in the Manage- other outcome measures (eg. and Dr Irrgang ing agility and perturbation training really helpful. ple. Dr Oddis. Dr Oddis. International Clinical Studies Including total knee arthroplasty. that single-limb stance time (a mea. 2010. A and perturbation training techniques 2008.16:137–162. 2010. ity and perturbation training tech. This study was approved by the University of have an additive effect on improving it would be helpful to determine Pittsburgh Institutional Review Board. Dr Wisniewski. risk factors.

WOMAC: a health status instrument for plasty: a randomized clinical trial. Fitzgerald GK. Campbell J.38:412– 421. ance exercise program appears to improve smith CH. The effectiveness of exercise therapy validation reapplication of the WOMAC in patients with osteoarthritis of the hip or osteoarthritis index. 1992. Rooks D. McMaster Universities Osteoarthritis Index) patients with knee osteoarthritis. Radiological of the lower extremities. J Clin Epidemiol. 1998. tion of osteoarthritis of the knee: Diagnos. 2005. instrument in the measurement of out. Janney J Rheumatol. Paul J. The CES-D scale: a self-report 1833–1840. Double blind randomized controlled trial of sodium meclofenamate (Meclomen) 25 van Baar ME. Clin J Pain. 12 Altman R. Measurement of pain. Surg Clin North Am. minimal clinically important difference. 1998. Irrgang JJ. measure of function of the knee. Axe MJ. 2004.90:1866 –1873.85:284 –289. 1993. parison of three scales. Stucki the American Rheumatism Association. Irrgang JJ. Measure- assessment of osteo-arthrosis. 2000. 2002.29:1039 –1049. Watson-Buchanan WW. Miller L. Celik A. anxiety. Validation study of function for patients with total knee arthro- 23 Beck AT. Phys Ther. 385– 401. ment of health status: ascertaining the Dis. osteoarthritis. Assessment arthritis of the knee: a randomized con- and reporting of osteoarthritis. 1977. 21 Jaeschke R. Relationships of fear. Gil AB.Agility and Perturbation Training Techniques in Exercise Therapy 10 Fitzgerald GK. matol. Snyder-Mackler L. Lachman M. Steer RA. Kean WF. Singer J. gerald GK. Appl Psychol Meas. Guyatt GH. classifica. Geron. Agility and perturbation train. tic and Therapeutic Criteria Committee of 14:343–349. Dekker J. Melzack R.46:153–162. Wisnewski S. 1999. G. Oostendorp RA. Phys Ther. 1998. 372–382. a randomized clinical trial. Wainner RS. J Consult relevant outcomes to antirheumatic drug Clin Psych.11:303–310. Michel BA. Physical Joint Surg Am. Ann Rheum 2002. J Bone 28 Fransen M. Smith KW.1: 16 Bellamy N. 466 f Physical Therapy Volume 91 Number 4 April 2011 . J Rheumatol. Effects of kinesthesia and balance exer- Home-based resistance training: predictors cises in knee osteoarthritis. Asch E. therapy in patients with osteoarthritis of the hip or knee. Aydin R. 1989. Phys Med Rehabil. matol. et al. 20 Katz J. Brown G. 28:156 –164. therapy is effective for patients with osteo- opment of criteria for the classification 19 Jensen MP. J Rheu- 19:153–159. 11 Fitzgerald GK. Piva SR. Get up and go test in patients with knee tologist. 1957. knee. Childs JD. 26 Scopaz KA. J Rheumatol. J Rheumatol. J Clin Rheu- 22 Piva SR. A bal- 15 Bellamy N. trolled clinical trial. Buchanan WW. CA. of pain during medical procedures: a com. and diclofenac sodium (Voltaren): post et al.90:880 – 894. 17 Hawker G. 1988. Gold.22:1193–1196.80:1132–1145. depression scale for research in a general population. 31 Piva SR. Edmonds J. specific (WOMAC) (Western Ontario and and depression with physical function in itation programs for physically active indi. Almeida GJ. Phys Ther. ety: psychometric properties. 1988. of participation and adherence. et al. 27 Washburn RA. The Physical Activity Scale for the ing for a physically active individual with 18 Irrgang JJ. son of a generic (SF-36) and disease. et al. Elderly (PASE): development and evalua- knee osteoarthritis. 14 Jette AM.79:231–252. Devel. 1986. Development of a patient-reported tion. 30 Diracoglu D.15: 24 Radloff LS. Compari. Ridge TM. Baskent A. Bloch D. Aeschlimann A. J Rheumatol. 2001. Arch viduals. et al. 2009. 29 Angst F. Epstein N. Minimal clinically important rehabilita- Arthritis Rheum. comes after knee replacement surgery.82: et al.80:128 –140.56:893– 897. Snyder-Mackler L. Control Clin Trials.16:494 –502. operative anterior cruciate ligament rehabil. 1998. Melfi C. Jette AM. 1995. Lawrence JS. tion effects in patients with osteoarthritis 13 Kellgren JH. Crosbie J. An inventory for measuring clinical anxi- measuring clincically important patient 2010. Fitz- The efficacy of perturbation training in non.10:407– 415. et al. et al. Fisher LD.25:2432–2439. Arch Phys Med Rehabil.29:131–138.

each of 30 seconds’ duration. then lower the limb back to the table. a towel. the hip and knee comfortably flexed and the foot resting comfortably on the floor. on the other limb. The subject holds the fully extended position for 3–5 seconds. Performed on wall. on the other limb. a 1-pound cuff weight is added. progressed by adding 1 pound when the subject can do 30 repetitions at the current resistance. The contralateral knee is flexed so that the foot is resting comfortably in a foot are given the option of doing 30 consecutive repetitions flat position on the table. The subject is of doing 30 consecutive repetitions or 3 sets of 10 instructed to hold the contraction for 3–5 seconds. discomfort is felt by the subject in the quadriceps. The Exercise is progressed from 10 to 30 repetitions. keeping the foot flat and maintained in slight supination and keeping the knee extended. The knee of the stretching both limbs. The exercise should be repeated on the other limb. Resistance is repeated on the other limb. instructed to hold the contraction for 3–5 seconds. Range can be increased during 30-second limb is placed in 90° of flexion. each of 30 seconds’ duration. Subject attempts to extend and flex knees in a range of motion from 0° to 45° of flexion against the to perform 3 sets of 10 repetitions at this resistance. Resistance is progressed by adding 1 pound when the subject can do 30 repetitions at the current resistance. Hamstring Stretching: The therapist stabilizes the contralateral limb on the plinth 2 repetitions. according to their preference. a 1- pound cuff weight is added. The subject is instructed to push against the force pad of the extension visit). The 2 repetitions. Range can be increased during 30-second foot is placed flat on the floor. as tolerated. the hip is extended. then the therapist extends the hip until a stretch period if subject reports stretch discomfort has decreased. resistance. The contralateral limb rests on the floor for stability. The therapist instructs the subject to raise the exercise limb or 3 sets of 10 repetitions. Subjects are given the option of doing device as vigorously as possible without reproducing pain symptoms. Quadriceps setting: The subject is positioned in long sitting with the knee Exercise is progressed from 10 contractions to 30 extended. Single-Limb Seated Leg Press: The subject is positioned in sitting on a leg press A resistance equivalent to 70% of the 1 repetition machine with the exercise limb fixed to the foot platform. Prone Hip Extensions: The subject is positioned prone on the treatment table. The exercise should be repeated according to their preference. The exercise should be repeated on the other limb. Therapist instructs the subject to isometrically contract the quadriceps contractions. A new 1 repetition maximum should be established every 2 weeks (every 4th visit). (Continued) April 2011 Volume 91 Number 4 Physical Therapy f 467 . until a stretch discomfort is felt in the calf muscles. The subject slowly leans forward toward the wall. Subjects table. Range can be increased during 30-second stretch discomfort is felt by the subject in the hamstrings while keeping the knee in period if subject reports stretch discomfort has decreased. Prone Quadriceps Stretching: The subject lies prone on the treatment table. Subjects therapist instructs the subject to raise the exercise limb with the knee maintained in are given the option of doing 30 consecutive repetitions or full extension as high as possible. then lower the limb back to the table. The minimum resistance is then 70% of the newly established 1 repetition maximum. each of 30 seconds’ duration. resistance should be advanced 1 plate (4. full extension. Performed on therapist stabilizes the contralateral limb on the plinth. A belt. The therapist then instructs the subject to extend the knee by sliding the foot along the treatment table toward the end of the table. The exercise 3 sets of 10 repetitions. the knee is extended. The of doing 30 consecutive repetitions or 3 sets of 10 subject holds the flexed position for 3–5 seconds. When the subject can perform 3 sets of 10 repetitions. Supine Straight Leg Raises: The subject is positioned supine on the treatment Exercise is progressed from 10 to 30 repetitions. Subjects are given the option far as possible by sliding the foot along the treatment table toward the pelvis. flexing the hip until a both limbs. The subject is 30 consecutive repetitions or 3 sets of 10 repetitions. according to their preference. For the limb being stretched. The exercise should be repeated repetitions. with the knee maintained in full extension to the height of the contralateral flexed When subject can do 30 repetitions without added knee position. Seated Knee Extension Isometrics: The subject is seated on a leg extension Exercise is progressed from 10 contractions to 30 exercise device with the knee in a comfortable flexed position between 90° and 60° contractions as soon as possible (by the 3rd treatment of flexion. Additional plate is added when 3 sets of 10 repetitions are achieved with current resistance. The exercise should be repeated on the other limb. The range can be increased during the 30-second period if the subject reports that the stretch discomfort has decreased. The subject is instructed maximum should be used for training. according to their preference. The exercise should be weight. Long-Sitting Knee Flexion and Extension: The subject is positioned in long Repetitions are progressed from a minimum of 10 to a sitting on the treatment table. When should be repeated on the other limb. Agility and Perturbation Training Techniques in Exercise Therapy Appendix. Subjects are given the option muscles bilaterally as vigorously as possible without reproducing pain. The exercise should be repeated on the other limb.54 kg). and the both limbs. subject can do 30 repetitions without added weight. used by the subject to assist with bending the knee. The exercise should be repeated on the other limb. Exercise Therapy Proceduresa Exercise Description Exercise Dosage/Progression Calf Stretching: Subject stands in front of wall with hands supporting body against the 2 repetitions. with period if subject reports stretch discomfort has decreased. Performed on and moves the stretching limb in a straight-leg-raise position. according to their preference. The therapist instructs the subject to flex the knee as maximum of 30 repetitions. or a strap may be repetitions.

according to their preference. The width of steps and the speed of steps are progressed approximately 10–20 ft. Therapist directs the subject to either walk forward. flexed to 90°. Continued Exercise Description Exercise Dosage/Progression Standing Hamstring Curls With Cuff Weights: A 1-kg cuff weight is wrapped Subject attempts to perform 3 sets of 10 repetitions. Changes in direction are cued randomly by the therapist. The exercise should be repeated on the other limb. The subject walks forward to first marker. Shuttle Walking: Plastic pylon markers will be placed at distances of 5. swelling. repeating 2 times in each direction for a total of 4 times. Braiding Activities: Subject combines front and back crossover steps while moving The activity is progressed by increasing the width of steps laterally (walking carioca). the exercise leg even with the thigh from the support leg. Tiltboard Balance Training: The subject stands on a tiltboard with both feet on The difficulty of the activity is progressed by adding ball the board. weight. moving right to left and then left to right. returns to 10-ft marker walking backward. The subject is instructed to raise up on the toes as high as possible. Therapist perturbs position. alternating legs with each step. approximately 10–20 ft. The activity lasts approximately 30 seconds. backward. every 1–2 sessions. starting with 1 plate of resistance (4. at random. Agility and Perturbation Activities for the Experimental Group Side Stepping: Subject steps sideways. Begin with tandem will cross one leg in front of the other. and the speed of steps can be progressed every 1–2 sessions. or buckling. or on diagonal by cueing the subject with hand signals. as tolerated. Multiple Change in Direction During Walking on Therapist Command: Duration of exercise bout is approximately 30 seconds. progressing to ball catching with therapist perturbing subject’s balance while standing on foam and progressing to single-leg support if tolerated without knee pain. both feet on the ground. the walking speed should be increased. and then left to right. (Continued) 468 f Physical Therapy Volume 91 Number 4 April 2011 . then return to the foot flat position. Keeping the thigh on When subjects can perform all 3 sets of 10 repetitions. The subject faces a wall or door. The therapist perturbs the tiltboard in forward and backward and side-to. backward. subject will be moving right to left and the speed of steps every 1–2 sessions. is able to tolerate this position without pain. alternating legs with each step. then finishes by walking to 15-ft marker. catching during the perturbations and progressing to side directions for approximately 30 seconds each. the exercise leg knee is another kilogram of resistance is added. Therapist attempts to perturb patient balance in random The difficulty is progressed as the subject improves by fashion. around the subject’s ankle. Front and Back Crossover Steps During Forward Ambulation: The subject Two repetitions are performed. Rollerboard and Platform Perturbations: The subject stands with one limb on a The activity may begin with subject in the semi-seated stationary platform and the other limb on a rollerboard. and the subject attempts to resist the difficulty doing the activity in full standing. The activity is perturbations. The activity is repeated by progressed to the full standing position when the subject changing the limbs on the platform and the rollerboard. The resistance should be advanced 1 plate when the subject can perform 30 repetitions. The subject will then walk backward to the start the subject’s performance improves. then returns to start by walking and the speed of steps every 1–2 sessions. at 1–5 minutes duration and progressing to 15 minutes. Treadmill Walking: Subject walks on a treadmill at a self-selected pace beginning When the subject reaches 15 minutes on the treadmill. The subject then walks to 15-ft marker. single-limb support perturbations if the subject tolerates single-limb weight bearing without knee pain. 10. or buckling. swelling. while walking crossover steps and progress to full crossover steps when forward approximately 10–20 ft. as tolerated. Subjects are given the option of doing 30 consecutive repetitions or 3 sets of 10 repetitions. The width of steps position while crossing one leg behind the other.54 kg). Standing Calf Raises: The subject is positioned in standing with both feet flat on When the subject can perform 30 repetitions with body the floor. then slowly lowered back to the start position. the exercise is performed on a calf machine. The activity is initiated on a level surface and progressed to side stepping over low obstacles when subject performs side stepping on level surfaces without difficulty. Subject then walks to 10-ft marker forward. Double-Leg Foam Balance Activity: Subject stands on a soft foam surface with The duration of the activity is approximately 30 seconds. holding for 1–2 seconds.Agility and Perturbation Training Techniques in Exercise Therapy Appendix. and The activity is progressed by increasing the width of steps 15 ft. with hips resting on plinth if the subject has rollerboard in multiple directions. During each activity. repeating 2 times in each direction for a total of 4 times. sideways. then returns to 5-ft marker walking backward.

et al.61:951–957. They perform single-leg standing balance. they perform all agility training with the exception of the multiple change in direction during walking on therapist command activity. The modifications are as follows: For the Standard Exercise Group’s Home Program: Wall squats are substituted for the seated leg press. Gil AB. Continued Home Exercise Program Subjects are encouraged to perform their exercises independently at home at least 2 times per week. Arthritis Rheum. They also do not perform tiltboard and rollerboard activities. Isometric knee extensions are performed against heavy resistance elastic bands that are secured to a chair. For the Agility and Perturbation Group’s Home Program: Subjects in the agility and perturbation group perform all of the activities in the standard exercise group’s home program activities. The program is essentially the same. with some modifications for home. 2009. a Reprinted with permission and with modifications from: Scopaz KA. April 2011 Volume 91 Number 4 Physical Therapy f 469 . In addition. Agility and Perturbation Training Techniques in Exercise Therapy Appendix. Piva SR. Effect of baseline quadriceps activation on changes in quadriceps strength after exercise therapy in subjects with knee osteoarthritis.

Reproduced with permission of the copyright owner. . Further reproduction prohibited without permission.