We recruited 17 older (mean age, 68 years; range, 55–80 years), moderately obese (mean body mass index [kg/m2
], 32.9; range, 25.9–39.7) individuals who underwent either a unilateral (n = 10) or bilateral (n = 7) TKA. All patients undergoing TKA who were recruited constituted a sample of convenience from an orthopaedic surgeon’s (CP) list of followup patients who were at least 12 months post-TKA (mean, 21 months; range, 12–53 months) and were medically cleared by their physician for physical exercise and lived in the Salt Lake valley region. Any previous revision to a TKA, clinical signs of rheumatoid arthritis, a progressive neurologic disorder, or a previous cerebrovascular incident constituted reasons for excluding subjects from participating. As well, any subject currently participating in a regular (two to three times per week) resistance exercise program was also excluded. There were 13 women and four men enrolled in the study. The participants’ pain, activity, recreation level, and health-related quality of life are similar to that previously reported (Table ) [21
]. The University of Utah’s Institutional Review Board approved this study. All subjects were informed and provided their written consent for participation in this study.
All TKA procedures were performed through a mini-medial parapatellar arthrotomy with minimal patella eversion [15
]. A conventional tricompartmental replacement was performed in all cases with the Biomet Vanguard knee system (Biomet, Warsaw, IN). After TKA, all subjects underwent inpatient care and home health physical therapy visits for the first 2 weeks after surgery. Thereafter, guidelines for outpatient physical therapy and independent exercises were provided to the subjects [30
]. Outpatient physical therapy sessions of 1 hour two to three times per week for 4 to 6 weeks were suggested. This included exercises (one to three sets of 10–15 repetitions) in seated, supine, and standing positions for knee range of motion; and neuromuscular electrical stimulation to augment quadriceps muscle activation, gait training, and walking in an obstacle course. A 10- to 15-minute stationary bicycling session was also suggested.
We attempted to match subjects by gender and whether they had received a unilateral or bilateral TKA, although with an odd number of bilateral TKA recipients (n = 7), the matched grouping was uneven and one group included an additional subject who had bilateral TKAs. First, the group of 17 patients was divided into seven pairs that were matched by gender and by whether the TKA was unilateral or bilateral plus one trio that consisted of one bilateral patient and two unilateral patients. Then a patient in each pair was randomly assigned to one of the two treatments, and within the trio, the bilateral patient was assigned to the traditional (TRAD) group (n = 8) and the two unilateral patients were assigned to the eccentric (ECC) lower extremity resistance exercise group (n = 9) for 12 weeks of rehabilitation. All participants were tested by one investigator (WM) before and after 12 weeks of training.
Both groups performed their respective lower extremity resistance exercise program for 30 minutes per session, 3 days per week, for 12 weeks. Compliance with the exercise program was based on the percentage of exercise sessions completed out of a total of 36 sessions over the 12-week period. Individuals were required to complete a minimum of 80% of the sessions to be included in the analyses. Before each training session, a 5- to 10-minute warmup on a standard cycle ergometer was performed. The TRAD group then performed four lower extremity resistance exercises (leg press, leg extension, leg curl, and calf raise) on weight training machines at 70% of their one-repetition maximum for three sets of 10 to 12 repetitions. Every other week, one-repetition maximum was reassessed and the exercise prescription was continued at a new resistance level commensurate to 70% of one-repetition maximum. Other traditional physical therapy exercises, ie, 4-inch stepups and wall squats, were incorporated into the exercise program using body weight as resistance at three sets of 10 to 12 repetitions.
The ECC group performed lower extremity resistance exercises exclusively on a recumbent eccentric resistance exercise stepper device described previously [5
]. The recumbent eccentric stepper is powered by a 3-horsepower motor that drives the foot pedals in a “backward” direction, ie, toward the individual. Eccentric muscle contractions occurred when the individual attempted to resist this motion by pushing on the pedals (with verbal instructions to “try to slow down the pedals”) as they moved toward them. Because the magnitude of the force produced by the stepper exceeds that of the individual, the pedals continue to move toward the participant at a constant velocity, resulting in eccentric contractions of the knee and hip extensors, including the quadriceps muscles (Fig. ).
Fig. 1 High muscle forces are generated on an eccentric stepper powered by a 3-horsepower motor that drives the pedals. As the pedals move toward the participant (largest arrow), the rider resists by applying force to the pedals (arrow at foot level). Because (more ...)
ECC subjects began with a 5-minute session on the stepper and progressed to a maximum of 20 minutes over the next 3 to 4 weeks at a self-selected range of 15 to 25 rotations per minute. The progression of the eccentric exercise work rate was dictated by the exercise protocol and determined as a function of the rating of perceived exertion based on the perceived exertion scale of Noble et al. [26
]. A “target” workload, at a constrained rating of perceived exertion level, was visible to the ECC group participant on a computer monitor and the goal was to achieve ever-increasing total amounts of negative work per session despite the constrained temporal and exertion levels. In the ECC group, the protocol dictated that the perceived exertion during exercise would progress from a “very, very light” [26
] exertion level for 5 to 11 minutes during the first week. During Week 2 (Sessions 4 to 6), the participants’ perceived exertion increased to a “very light” exertion level and 3 minutes of training time was added to each session until a maximum 20 minutes of training was achieved. In Week 3, ECC training was increased to a “fairly light” exertion level for 20 minutes. In the last 8 weeks of training, the perceived exertion was between a “fairly light” and “somewhat hard” level with each session being 20 minutes in length. It is our experience that during any form of exercise, this population will exert to a level consistent with the perception of exercising “somewhat hard.” Therefore, once a rating of perceived exertion level of “somewhat hard” was achieved, participants were instructed to maintain that exertion level throughout the remaining weeks of training. Immediately after each training session, the total negative work in kilojoules (kJ) was recorded and averaged over the number of exercise sessions per week.
Both thighs of all patients underwent MRI 1 week before and 1 week after the 12 weeks of exercise to assess the muscle volume of the quadriceps as previously described [5
]. Participants were placed supine in the MRI unit with the legs relaxed. All scans were performed on one 1.5-Tesla whole-body MRI unit (Signa Lightning LX 8.4; General Electric Medical Systems, Milwaukee, WI). To establish the region of interest, a coronal fast spoiled gradient echo scout scan was used to identify the superior and inferior boundaries of the scans (the femoral head to the TKA knee system component parts representing the tibiofemoral joint line). Once the region of interest was established, axial T1-weighted images were acquired in the standard body coil using a fast spin echo sequence: 8-mm slice thickness, 15-mm interslice distance, and a 320 × 320 matrix. Depending on thigh length, the number of sections acquired ranged from 17 to 22. (Note: the identical number and location of slices were used in the comparison from pre- to posttraining.) The axial MRI images were then digitized and saved to compact disc for later analysis.
After electronic data transfer of images, muscle volume measurements and calculations were performed by use of custom-written image analysis software (MatLab; Mathworks, Inc, Natick, MA) on a desktop personal computer. For each image, the quadriceps muscle of interest, eg, vastus lateralis, vastus medialis, vastus intermedius, and rectus femoris (independent of skin, bone, and fat), were identified from the displayed images. One of the authors (WM) manually traced the muscle outlines using a computer mouse, allowing overall muscle volume to be automatically computed. Muscle volume was determined by summing the volumes from each slice (cross-sectional area × slice thickness) to give total volume as described by previous researchers [37
]. The same investigator, blinded to time point of the scan and slice location, performed measurements of individual participants’ quadriceps muscle before and after training. To establish intrainvestigator reliability of the muscle volume measurement, the same investigator performed two separate measurements of quadriceps muscle volume of 18 different images on six individuals. The average interclass correlation coefficient across the 18 images was 0.99. The validity of the volume measurement was determined by analysis of images obtained from a cadaveric thigh phantom that approximated the size of the quadriceps femoris muscle group. The volume of the phantom, measured by water displacement 5 hours after MRI scanning, was 100.7% of the MRI-determined value. There was a 0.012% difference between repeat volume displacement measurements of the phantom by the same investigator [5
Lower extremity knee extension strength was quantitatively assessed by unilateral maximal voluntary isometric force on a KinCom dynamometer (Chattanooga Inc, Hixon, TN). Previous research has supported the reliability of this measure [3
]. Both lower extremities were tested and these strength measures were assessed before and 2 to 5 days after the training interventions. Participants were seated and their knees were fixed at 75° of flexion. Before testing, participants practiced submaximal contractions at 50% and 75% of their maximal effort. One practice trial was then performed. After a brief rest period, three separate maximal contractions were performed, each held for 5 seconds with a 3-minute rest between trials. The outcome variable muscle torque was calculated as the peak torque of three trials. The order of testing (more affected versus less affected limb as determined by the subject’s perception of leg impairment) was randomized among subjects. Subjects were stabilized by chest and thigh straps and asked to fold their arms across their chest while performing these tests.
A battery of three reliable mobility tasks regularly used with older individuals and responsive to detecting change [10
] was used to determine the functional relevance of any muscle strength changes. All mobility measurements were performed by one investigator (WM). All participants underwent this series of tests before and after (within 2–5 days) 12 weeks of training. The timed up and go test is a timed (s) test in which participants begin in a seated position in an armchair and then rise, go forward 3 m, turn around, and sit back down. The timed up and go test can discriminate between older individuals who have fallen and those who have not fallen. The 6-minute walk test, a measure of the distance (m) an individual walks in 6 minutes [11
], was used to assess overall locomotor ability and locomotor fatigue. Participants were asked to cover as much distance as possible within 6 minutes. The stair ascent and descent time tests (s) [3
] were used to assess functional mobility and the use of concentric and eccentric lower extremity muscle force production abilities, respectively. Participants were asked to ascend and descend one flight of stairs under close or contact supervision as quickly and safely as possible. Time was recorded to the nearest 0.01 second from a verbal go signal to final foot placement on a standard flight of 10 stairs and the average of three trials was recorded. Previous research has supported the validity of these measures [29
] and reported quadriceps muscle strength as being positively correlated 12 months after TKA with faster times during the timed up and go and stairclimbing tests and greater distances covered during the 6-minute walk test [29
Because no previous studies have examined the practicality of eccentric exercise in older individuals after TKA, as control variables, we documented the ECC group participants’ subjective interpretation of whether the eccentric training induced any leg muscle pain and/or any knee pain. A 10-cm visual analog scale (0 cm = no pain, 10 cm = worst possible pain) for both the leg muscles and knee were filled out by the participant before each training session (to document residual pain from previous sessions).
Descriptive statistics were calculated for demographic variables and dependent measures. In the analysis of the dependent measures, the assumptions underlying parametric tests, normality, and homogeneity of variance were met. In accordance with the dual purpose of the study, we evaluated: (1) whether 12 weeks of resistance exercise, as part of a rehabilitation program, can induce changes in muscle size, strength, and mobility in older individuals 1 to 4 years after their TKA; and (2) whether the effect of the type of resistance exercise (ECC versus TRAD) influenced muscle and mobility outcomes. To determine within-group pre- to posttraining changes in muscle and mobility, we used a one-way repeated-measures analysis of variance. To determine between-group differences in the pre- to posttraining changes in muscle and mobility, we used a two-way repeated-measures analysis of variance with particular attention paid to the time by group interaction effects because a significant result would reflect a differing response of groups over time. Significant interactions (p < 0.05) were then assessed through a pairwise comparison test using a Bonferroni correction. Data were analyzed with Sigma Stat Version 3.5 (Chicago, IL).