ClinicalTrials.gov identifier: NCT00078624.
Associations for change in physical and psychological factors and treatment response following exercise in knee osteoarthritis: An exploratory study†
Article first published online: 27 OCT 2012
Copyright © 2012 by the American College of Rheumatology
Arthritis Care & Research
Volume 64, Issue 11, pages 1673–1680, November 2012
How to Cite
Fitzgerald, G. K., White, D. K. and Piva, S. R. (2012), Associations for change in physical and psychological factors and treatment response following exercise in knee osteoarthritis: An exploratory study. Arthritis Care Res, 64: 1673–1680. doi: 10.1002/acr.21751
- Issue published online: 27 OCT 2012
- Article first published online: 27 OCT 2012
- Accepted manuscript online: 5 JUN 2012 10:35AM EST
- Manuscript Accepted: 21 MAY 2012
- Manuscript Received: 13 DEC 2011
- National Institute of Arthritis and Musculoskeletal and Skin Diseases. Grant Number: 1-R01-AR048760
Understanding how changes in physical and psychological factors following therapeutic exercise are associated with treatment outcome could have important implications for refining rehabilitation programs for knee osteoarthritis (OA). The objective of this study was to examine the association of changes in these factors with changes in pain and function after an exercise program for people with knee OA.
In total, 152 people with knee OA completed an exercise program consisting of lower extremity strengthening, stretching, range of motion, balance and agility, and aerobic exercises. The change from baseline to the 2-month followup was calculated for physical and psychological factors, including self-reported knee instability, quadriceps strength, knee and ankle range of motion, lower extremity muscle flexibility, fear of physical activity, anxiety, and depressive symptoms. Treatment response was defined as a minimum of a 20% improvement from baseline in both a numerical knee pain rating scale and the Western Ontario and McMaster Universities Osteoarthritis Index physical function scale. The association of each factor with treatment response was examined with logistic regression mutually adjusted for age, sex, body mass index, radiographic severity, and exercise group.
Change in self-reported knee instability (odds ratio [OR] 1.67 [95% confidence interval (95% CI) 1.13–2.47]) and fear of physical activity (OR 0.93 [95% CI 0.88–1.00]) were the only 2 factors that were significantly associated with treatment response after adjustment for covariates.
Improvement in knee instability and fear of physical activity were associated with increased odds of a positive treatment response following therapeutic exercise in subjects with knee OA.
Improvements in pain and physical function are important clinical outcomes for people with knee osteoarthritis (OA). Exercise therapy is an important part of clinical management for people with knee OA (1–5). Although exercise therapy is helpful, recent systematic reviews indicate that the effects of exercise on pain and function yield small to moderate effects for people with knee OA (6, 7). Exercise therapy programs for knee OA have traditionally been designed to address impairments in physical factors associated with knee OA, such as lower extremity joint motion deficits, muscle weakness, and reduced aerobic capacity (1–4). While it is commonly assumed that improving impairments in these physical factors will be associated with improvements in pain and function in response to therapeutic exercise, this assumption has not been widely examined.
There is also increasing evidence that psychological factors such as fear, anxiety, and depression have adverse effects on disability in people with knee OA (8–10). There is evidence that exercise and physical activity improve factors related to psychological distress (11, 12). It would seem reasonable that if exercise had a positive effect on fear, anxiety, and depression, then changes in these psychological factors in response to exercise could be associated with improved pain and function in people with knee OA.
Understanding the associations between changes in physical and psychological factors following exercise therapy with changes in pain and function could have important implications for refining exercise therapy programs for people with knee OA. If improvement in certain factors was found to be more strongly associated with improvement in pain and function than others, these certain factors might be more likely to be the active ingredients for improving pain and function. Future refinement of exercise therapy programs might be more focused in addressing these factors than factors where the change had little association with change in pain and function. The refinement could lead to larger treatment effects from therapeutic exercise, or similar effects but with more time- and cost-efficient approaches (13).
Recently, we reported the results of a randomized trial comparing exercise approaches for reducing pain and improving physical function in subjects with knee OA (14). The intervention groups included a standard impairment-based exercise group that performed lower extremity strengthening, range of motion (ROM), and flexibility exercises, and an experiment group that received agility and balance training in combination with the same standard impairment-based exercise program. Details of the interventions have been described in previous reports (14, 15). Although both groups demonstrated improvements in pain and function from baseline to posttreatment followup, there was no overall difference between the groups at followup.
During the randomized trial, we collected data on changes in several factors that could potentially be associated with clinical outcome, which included lower extremity ROM and flexibility, quadriceps strength, self-reported knee instability, and psychological variables such as fear of physical activity, anxiety, and depression. The purpose of the current study was to explore whether the changes in these factors were indeed associated with treatment outcome in our trial. Specifically, we examined whether changes in these factors predicted improvements in pain and function. We hypothesized that subjects who improved in the physical and psychological factors would be more likely to have improvements in pain and function.
Significance & Innovations
Although exercise therapy programs are designed to mitigate impairments in physical factors associated with knee osteoarthritis (OA) with the intent of improving pain and function, little is known about the association between changes in impairments and clinical outcome (improvements in pain and function) following exercise therapy. Identifying the changes in impairments that are most strongly associated with treatment response could lead to a refinement of exercise therapy programs to target the more active ingredients in improving pain and function.
This study examined the association between changes in physical factors (such as quadriceps muscle strength, knee and ankle range of motion, lower extremity muscle flexibility, and self-reported knee instability) and psychological factors (such as fear of physical activity, anxiety, and depression) and the response to treatment with regard to pain and function following an exercise therapy program.
The results indicate that improvement in self-reported knee instability and fear of physical activity are associated with positive treatment outcome and suggest that supplementary interventions to target knee instability and or fear of physical activity in conjunction with exercise therapy may enhance the overall effectiveness of treatment in people with knee OA who have these problems.
SUBJECTS AND METHODS
The data reported here were taken from a randomized clinical trial of exercise therapy for subjects with knee OA (14). The current study used baseline and 2-month followup data from the randomized trial.
The study was conducted at the University of Pittsburgh Medical Center, Center for Sports Medicine in Pittsburgh, Pennsylvania. The subjects were recruited from the Pittsburgh metropolitan area through physicians' offices, community flyers, newspaper advertisements, and the University of Pittsburgh Arthritis Institute Registry.
In total, 152 subjects were eligible. Subjects were eligible if they had complete clinical outcome data at both the baseline and the 2-month followup time points. Subjects were included if they were ≥40 years of age and met the 1986 American College of Rheumatology clinical criteria for knee OA (16). Exclusion criteria were the presence of any condition that would place the subject at risk for injury during exercise (e.g., required assistive devices for ambulation, a history of ≥2 falls in the previous year), a history of total knee arthroplasty, uncontrolled hypertension, cardiovascular disease, neurologic disorders affecting lower extremity function (e.g., stroke, peripheral neuropathy), and vision problems affecting the performance of basic mobility tasks. All subjects signed an informed consent form approved by the University of Pittsburgh Institutional Review Board before participation in the study.
Subjects were randomized to 1 of 2 types of exercise therapies, as indicated above. Physical therapists supervised 12 sessions of therapy over the course of 6–8 weeks.
Measurement of clinical outcome.
Clinical outcome was categorized as responder or nonresponder based on the changes from baseline in the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) physical function subscale and an 11-point numerical knee pain rating scale (NKPRS). The WOMAC physical function subscale is a disease-specific measure of function consisting of 17 questions about physical function, where individual items are scored from 0 = no difficulty to 4 = extreme difficulty. Reliability, validity, and responsiveness of the WOMAC have been established (17–19). Subjects rated their worst knee pain experienced in the past 24 hours prior to testing using the NKPRS, where 0 = no pain and 10= the worst pain imaginable. Numerical rating scales were found to be reliable and valid for measuring clinical pain (20, 21).
For classification as a treatment responder, subjects had to have 20% improvement from baseline to the 2-month followup in both the WOMAC physical function subscale and the NKPRS. We believed that requiring at least moderate improvement in both pain and function would enhance the likelihood that the subjects we classified as responders had a successful outcome of treatment. A 20% improvement in pain or function from baseline is considered moderate improvement (22).
Measurement of physical and psychological factors.
Self-reported knee instability.
Subjects rated their severity of knee instability on a 0–5-point numerical scale in response to the query, “To what degree does giving way, buckling, or shifting of the knee affect your level of daily activity?” where 5 = “I do not have the symptom,” 4 = “I have the symptom but it does not affect daily activity,” 3 = “the symptom affects my activity slightly,” 2 = “the symptom affects my activity moderately,” 1 = “the symptom affects my activity severely,” and 0 = “the symptom prevents me from all daily activities.” This self-report rating of knee instability was taken from the Knee Outcome Survey Activities of Daily Living Scale (23). The test–retest reliability of this self-report rating of knee instability is good (intraclass correlation coefficient [ICC2,1] 0.72).
Maximum voluntary isometric quadriceps torque was measured using a Biodex System 3 dynamometer. A minimum of 3 trials and a maximum of 6 trials were performed. After 3 trials, when a trial had a maximum torque less than the previous trial, the strength testing was concluded. The highest maximum torque from all trials was recorded as the quadriceps strength score. This procedure has been shown to yield reliable quadriceps torque measurements in our laboratory (ICC2,1 0.96).
Knee flexion and extension ROM.
Knee flexion and extension ROM were measured in degrees using standard goniometric procedures. The subject was supine for these measurements.
Ankle dorsiflexion ROM.
Ankle dorsiflexion ROM was measured using standard goniometric procedures. This measurement was taken with the patient prone and the knee positioned at 90° of knee flexion to eliminate the effect of the gastrocnemius muscle on ankle motion and also at 0° of knee flexion to maximize the gastrocnemius muscle effect on ankle motion.
Hip flexibility (Thomas test ).
The subject was supine with the nontested hip flexed so the sacrum was flat on the table. The test extremity's thigh was brought to the table surface and the lower extremity was dangled over the edge of the table. Hip flexibility was normal if the thigh rested flat on the table surface. If the thigh did not rest on the table surface, hip flexion in this position was measured as the degree of hip flexion contracture.
Rectus femoris flexibility (modified Thomas test ).
The subject was positioned as described for the Thomas test. The test knee was permitted to flex to end range. Knee flexion was measured in this position as the amount of rectus femoris flexibility.
First, the popliteal angle was measured with the subject supine and the hip flexed to 90°, then extending the knee as far as possible while recording the knee flexion in this position. Next, the previously obtained value for knee extension ROM with the patient supine was subtracted from the popliteal angle to provide the final measure of hamstring flexibility. Smaller numbers represent better hamstring flexibility.
Gastrocnemius flexibility was determined by subtracting ankle dorsiflexion measured with the knee at 90° of knee flexion from ankle dorsiflexion measured with the knee fully extended. Smaller numbers represent better gastrocnemius flexibility.
Fear of physical activity.
Fear-avoidance beliefs were measured using the physical activity scale of the Fear-Avoidance Beliefs Questionnaire (FABQ), which was originally developed for low back pain (25). A modified version of the FABQ physical activity scale has been used by van Baar et al to assess fear-avoidance beliefs in people with knee pathology (26). For our study, the FABQ physical activity scale was modified from the low back pain version by replacing the word “back” with “knee,” and replacing the physical activity examples “bending, lifting, walking, and driving” with “running, walking, kneeling, and driving.” The FABQ physical activity scale quantifies the level of fear associated with physical activity, including beliefs about the association of physical activity with knee pain and harm. The scale consists of 4 items, each scored from 0–6, with a maximum of 24. Higher scores represent greater fear-avoidance beliefs. The Cronbach's α of the FABQ physical activity scale in our sample of patients with knee OA was 0.75, indicating adequate internal consistency.
Depression was measured using the Center for Epidemiologic Studies Depression Scale (CES-D) (29). The CES-D is a 20-item self-report depression scale, where each item is scored on a 0–3 scale with higher scores representing greater depression. The CES-D is a reliable, valid, and sensitive tool for detecting depression and changes in depression over time (29–31).
Prior to baseline measurement, subjects indicated which knee was most symptomatic, and this knee was used as the index knee for all knee-specific measures in the analyses (radiographic severity, quadriceps strength, knee and ankle ROM, and lower extremity muscle flexibility). In order to describe the sample, descriptive statistics were calculated for demographic variables with respect to treatment response. The change score from baseline to the 2-month followup was calculated for each physical and psychological factor by subtracting the baseline score from the 2-month followup score. Means and SDs for these change scores were calculated for treatment responders and nonresponders. Independent t-tests were performed to examine group differences in change scores.
The associations of change in each physical and psychological factor with treatment response were examined in separate unadjusted logistic regression models. Those univariate associations significant at the P value of 0.10 or less level were included in the multivariate logistic regression adjusted for other physical or psychological factors meeting this criterion along with age, sex, body mass index (BMI), treatment group, and radiographic severity of knee OA. Regression analyses were referenced such that higher odds ratios indicate a higher likelihood of treatment response. The significance level for the multivariate logistic regression was a P value less than or equal to 0.05.
Because our primary treatment response variable was derived from clinically relevant change in both pain and function, it is possible that some physical and psychological factors might be more associated with pain but not with function, or vice versa. To better explain the associations, we also performed the above regressions using only change in pain or only change in function as the outcome variable. Again, subjects were considered responders if they achieved at least a 20% change from baseline in either pain or function.
The demographic information of the subjects is shown in Table 1. The demographic characteristics between the responders and nonresponders were similar. Information on factor change variables between treatment responders and nonresponders is shown in Table 2. We were unable to obtain change measures on some subjects for each of the physical factors, as is indicated by the varying sample sizes in Table 2. This was because some subjects did not show up for their 2-month physical examination, and some subjects refused to perform some measures on the followup examination. We had complete data for all 152 subjects on the self-report measures because subjects were given the option to mail in their followup surveys even if they could not come in for the 2-month followup physical examination. Group differences were observed for changes in self-reported knee instability and fear of physical activity, with responders demonstrating greater improvement on these variables. Change in quadriceps torque appeared to be greater for the responder group; however, this difference was not statistically significant. All other impairment change variables were not significantly different between the groups.
|Nonresponders (n = 99)||Responders (n = 53)|
|Age, mean ± SD years||65.06 ± 8.43||62.57 ± 9.04|
|Body mass index, mean ± SD kg/m2||30.10 ± 5.82||30.31 ± 6.90|
|Sex, no. (%)|
|Male||32 (32.32)||21 (39.62)|
|Female||67 (67.68)||32 (60.38)|
|Radiographic severity, no. (%)*|
|1||1 (1.01)||1 (1.89)|
|2||14 (14.14)||7 (13.21)|
|3||45 (45.45)||29 (54.72)|
|4||39 (39.39)||15 (28.30)|
|Mean ± SD||Range||No.||Mean ± SD||Range||No.|
|WOMAC physical function subscale||−1.89 ± 6.98||−36.0 to 14.0||99||−10.87 ± 6.96||−32.0 to −1.0||53||< 0.01|
|Numerical knee pain rating scale||−1.34 ± 4.30||−15.0 to 7.0||99||−1.90 ± 4.34||−12.0 to 6.0||53||0.48|
|Knee instability||0.05 ± 1.20||−3.0 to 5.0||99||0.62 ± 0.88||−1.0 to 3.0||53||< 0.01|
|Quadriceps strength||0.76 ± 19.06||−72.0 to 50.0||92||7.21 ± 20.59||−32.0 to 98.0||52||0.06|
|Knee flexion ROM||2.82 ± 1.73||−10.0 to 14.0||94||3.31 ± 1.87||−7.5 to 15.0||53||0.59|
|Knee extension ROM||0.94 ± 5.18||−11.5 to 28.0||94||1.76 ± 5.98||−14.5 to 22.5||53||0.38|
|Knee total ROM||3.76 ± 7.10||−12.0 to 27.0||94||5.08 ± 7.00||−8.5 to 24.0||53||0.28|
|Ankle dorsiflexion ROM||0.22 ± 5.03||−13.0 to 15.5||88||0.29 ± 5.27||−9.0 to 11.0||52||0.94|
|Hip flexor flexibility||−3.53 ± 4.47||−19.0 to 5.0||90||−2.93 ± 4.27||−14.0 to 5.5||52||0.43|
|Rectus femoris flexibility||8.71 ± 12.36||−23.5 to 41.0||92||10.75 ± 12.04||−24.0 to 37.5||53||0.34|
|Hamstring flexibility||2.85 ± 14.69||−26.5 to 38.5||94||5.68 ± 12.67||−22.0 to 36.5||53||0.24|
|Gastrocnemius flexibility||−1.80 ± 6.83||−23.0 to 21.5||86||−1.86 ± 5.81||−15.0 to 9.0||52||0.96|
|Fear of physical activity†||−0.28 ± 6.41||−18.0 to 15.0||99||−2.91 ± 5.57||−22.0 to 8.0||53||0.01|
|Anxiety‡||−0.75 ± 4.18||−20.0 to 10.0||99||−1.49 ± 3.33||−9.0 to 8.0||53||0.26|
|Depression§||−1.30 ± 5.94||−24.0 to 13.0||99||−0.43 ± 4.39||−13.0 to 12.0||53||0.35|
Table 3 shows the summary of the results from the regression analyses. When the combined change in pain and function was used as the outcome, self-reported knee instability and fear of physical activity were significant predictors of treatment outcome in both the unadjusted and adjusted models. After adjusting for covariates, for each 1-level improvement in knee instability rating, subjects had 1.7 times the odds of a positive treatment response. For each level of reduction in fear of physical activity, subjects had a 7% increase in the odds of a positive treatment response. Quadriceps torque had a small association with treatment response, but this association was not significant (P = 0.06). None of the other factors were associated with treatment response.
|Change variables||Pain and function (n = 53 responders)†||Function only (n = 84 responders)†||Pain only (n = 72 responders)†|
|Unadjusted OR (95% CI)||Adjusted OR (95% CI)‡||Unadjusted OR (95% CI)||Adjusted OR (95% CI)‡||Unadjusted OR (95% CI)||Adjusted OR (95% CI)‡|
|Self-reported knee instability||1.63 (1.16–2.29)§||1.67 (1.13–2.47)§||1.52 (1.10–2.10)¶||1.61 (1.13–2.29)§||1.54 (1.12–2.14)§||1.59 (1.13–2.23)§|
|Quadriceps torque||1.02 (0.99–1.04)||1.02 (0.99–1.04)||1.02 (1.00–1.03)||1.02 (1.00–1.04)||N/A#||N/A#|
|Fear of physical activity**||0.93 (0.88–0.99)¶||0.93 (0.88–1.00)¶||0.96 (0.91–1.01)||N/A#||N/A#||N/A#|
When the change in pain and function was analyzed separately, self-reported knee instability was the only factor that predicted treatment response. After adjusting for covariates, for each 1-level improvement in knee instability rating, subjects had 1.6 times the odds of a positive treatment response. No other factors were predictive of treatment outcome when change in pain and change in function were considered separately.
Our results indicate that subjects who improve in self-reported knee instability or reduce their fear of physical activity during the course of treatment are more likely to have a positive response to treatment in terms of pain and function. The findings may also suggest that if patients with knee OA are reporting problems with knee instability or fear of physical activity, perhaps their pain and function could be further enhanced by using supplementary interventions that specifically target these problems. One such treatment for addressing knee instability may be knee orthotics. Ramsey et al demonstrated that neutral knee bracing could significantly reduce self-reported knee instability in patients with knee OA (32). In our previous study, we believed that adding perturbation and agility training to knee OA rehabilitation programs might also be a method to improve dynamic knee stability (14, 33); however, we were unable to show that adding such interventions was superior to conventional impairment-based exercise in improving self-reported knee instability (14). In any case, our current results may suggest that exploring other methods for specifically addressing knee instability may enhance the overall effectiveness of exercise in subjects with knee OA who have knee instability.
Although we are unaware of any current treatment paradigms that specifically address fear of physical activity in people with knee OA, programs have been developed to address fear in patients with low back pain (34, 35). The interventions are designed to disconfirm beliefs that engaging in physical activities will harm the individual. This is done by educating patients on proper adaptive behaviors versus maladaptive behaviors in response to pain, encouraging subjects to focus attention on physical activity accomplishments rather than on the level of pain to gauge progress, instruction in proper ergonomic principles for performing various physical activities, and graded exposure to exercise and fear-producing physical activities in therapy to gradually reduce fear associated with the performance of these activities. This type of approach may also be beneficial for people with knee OA who are fearful of engaging in physical activity, but the research for this population is limited.
We were surprised that the change in other physical factors, such as quadriceps strength, knee motion, and lower extremity muscle flexibility, was not significantly associated with treatment response. It is known that these impairments are associated with physical disability in people with knee OA (36–39); however, changes in these variables were not associated with treatment response in our study. Several previous studies have reported group improvements in quadriceps strength that were concomitant with group improvements in pain and function measures (40–48). However, few have actually examined the direct relationship between change in strength and change in treatment outcome based on changes in pain and function.
While it seems logical that if a group changes in quadriceps strength and function simultaneously then these 2 factors should be related, it must be realized that group mean changes do not necessarily reflect variation in the individual changes in strength, pain, and function that would drive the relationship between these factors. Therefore, to establish a definitive relationship between change in strength and change in pain and function, this association should be examined directly. We found only 2 studies examining this relationship directly. In one study (45), the authors stated that increased knee extensor strength was associated with decreased pain and increased function based on linear regression, but they did not provide any information on the model, whether there was control of covariates, or a summary of the model results. In the other study (46), a simple Pearson's correlation coefficient was provided, indicating that improvements in knee extension strength were modestly correlated with improvements in function, but there was no control for covariates such as age, BMI, sex, or radiographic severity. In our study, the treatment responders seemed to have greater changes in strength compared to nonresponders, but this difference did not achieve statistical significance. Likewise, after controlling for demographic covariates, the association between changes in quadriceps strength and treatment outcome in our study approached statistical significance, but the magnitude of this association was also very small. Another consideration is that the measures of strength used to assess the relationship between quadriceps strength and function are maximum strength measures, yet maximum muscle torque output is not required to perform most daily functional tasks. This might also explain our lack of association between these 2 variables. Taken together, we believe that the degree to which change in quadriceps strength can influence change in self-reported pain and function is unclear at the present time.
It may be possible that for some people, engaging in an exercise program would improve their confidence that their knee will not buckle (improve knee stability). This improved confidence might also diminish their fear of performing previously difficult functional activities. The improved confidence resulting from reduced experiences of instability and fear may be a greater determinant in improving a patient's perception of pain and function than improvement in muscle strength and joint mobility. This could explain why these factors dominated over the other physical factors in predicting treatment outcome in our sample.
While we believe our results may suggest that incorporating treatments to target knee instability and fear (for people who have these problems) may improve the effects of exercise on pain and function, this suggestion assumes that the causal pathway is unidirectional and that change in instability and fear leads to change in pain and function. We must acknowledge that we do not know from our study that this causal direction is definitive. It is possible that change in pain and function may also drive the change in instability and fear. If this is the case, then it is possible that targeted interventions for instability and fear may not be successful in enhancing the effects of exercise on pain and function. However, we believe that our data do provide enough support to at least test our assumption by examining whether these targeted treatments may enhance the effect of exercise therapy on pain and function.
There are limitations to the results of our study. The measure of self-reported knee instability used in our study queried subjects on how knee instability affects their activities of daily living. Given the link of the symptom to activities of daily living in the question, there is a possibility that the association between change in self-reported knee instability and treatment outcome in our study may be an overestimate of the true association, because self-reported function was part of the treatment outcome measure. While we acknowledge this possibility, we do not believe this was a significant problem in our study. As indicated in Table 3, the association of change in knee instability with improvement in pain (the other treatment outcome measure) was identical to the association of change in knee instability with change in self-reported function. Our pain measure was not tied to specific functional tasks. Second, another study reported a similar association between self-reported knee instability and physical function measures when the question asked of the subjects did not link knee instability with activities of daily living (49). In this regard, we believe we are measuring a similar construct.
There are other impairments we did not measure in our study, such as aerobic capacity, balance, and deficits in self-efficacy, that could change as a result of therapeutic exercise intervention and might be associated with treatment outcome. We were also not able to assess how disease progression may interact with participation in exercise therapy over a long period of time and how this might affect treatment response. Our results are based on self-reported measures of treatment response and may not apply to performance-based measures of function. Likewise, our results only apply to therapeutic exercise programs similar to what were used in the current study. Our findings may not necessarily apply to rehabilitation programs that incorporate other types of therapeutic interventions. For example, rehabilitation programs that incorporate manual therapy techniques in conjunction with therapeutic exercise have yielded fairly large effect sizes for improving pain and function (50, 51). Manual therapy techniques address impairments in joint motion more aggressively than the interventions used in our therapeutic exercise program. Although improving joint mobility was not found to be significantly associated with treatment response in our study, it is possible that improving joint mobility could become significantly associated with outcome in programs that use manual therapy techniques.
Our results indicate that improvement in self-reported knee instability and fear of physical activity are associated with treatment response to the therapeutic exercise programs used in our study. This finding may suggest that methods to improve self-reported knee instability and fear of physical activity might further enhance the effects of therapeutic exercise for patients with knee OA who have these problems. Continued work is needed to confirm the associations between changes in physical and psychological factors and treatment response to therapeutic exercise for people with knee OA.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Fitzgerald had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Fitzgerald, White, Piva.
Acquisition of data. Fitzgerald, Piva.
Analysis and interpretation of data. Fitzgerald, White, Piva.
- 3Ottawa Panel. Ottawa Panel evidence-based clinical practice guidelines for therapeutic exercises and manual therapy in the management of osteoarthritis. Phys Ther 2005; 85: 907–71.
- 6Exercise for osteoarthritis of the hip or knee. Cochrane Database Syst Rev 2003; 3:CD004286., , .
- 23Development of a patient-reported measure of function of the knee. J Bone Joint Surg Am 1998; 80-A: 1132–45., , , , .
- 24Muscles: testing and function. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 1993., , .
- 28Use of the Beck Anxiety and Depression Inventories for primary care with medical outpatients. Assessment 1997; 4: 211–9., , , , .
- 29The CES-D scale: a self-report depression scale for research in a general population. Appl Psychol Meas 1977; 1: 385–401..
- 39Associations between physical examination and self-reported physical function in community dwelling older adults with knee pain. Phys Ther 2008; 88: 1–10., , , , .