The influence of pathology, pain, balance, and self-efficacy on function in women with osteoarthritis of the knee.

Background and Purpose. The determinants of function in people with osteoarthritis include both psychosocial and physiological variables. Studies that simultaneously integrate both domains are limited. The aim of this study was to determine the following in a sample of women with osteoarthritis (OA) of the knee: (1) the relationships among pathology (grade of OA of the knee), pain level, balance, and self-efficacy and (2) the relative effects of pathology, pain level, balance, and self-efficacy on function. Subjects. Fifty community-dwelling women, 50 to 84 years of age ([bar.X]=69.2, SD=8.8), with symptoms of OA of the knee participated. Methods. Radiographs, standardized questionnaires (the Western Ontario and McMaster Universities Osteoarthritis Index, the Arthritis Self-Efficacy Scale, the Functional Reach Test, and timed performance tests were used to quantify the variables. Bivariate analyses and stepwise multiple regression modeling with analysis of variance calculations for beta weight testing were used in data analysis. Results. In regression analysis, functional self-efficacy and balance accounted for 42% of the variance in physical performance of function, whereas functional self-efficacy and pain accounted for 74% of the variance in self-report of functional difficulty. Discussion and Conclusion. Functional self-efficacy is an important factor affecting the functional performance outcome for people with OA of the knee. Suggestions are given to address self-efficacy in health care management. [Harrison AL. The influence of pathology, pain, balance, and self-efficacy on function in women with osteoarthritis of the knee. Phys Ther. 2004;84:822-831.]

Key Words: Aging, Knee, Osteoarthritis, Pain.

**********

Managing acute pathology often relies on the relationship the clinician believes exists between signs and symptoms and disease diagnosis. Manage the pathology, the theory suggests, and the signs and symptoms will diminish. In a study sample of older adults, however, Fried et al (1) found that this model fit fewer than half of the people in a sample of older patients with varying pathologies examined in a geriatric medicine clinic and noted that older adults are more likely to have chronic pathologies. People with chronic pathologies often have fluctuating symptoms related to exacerbations and remissions and the lack of an identifiable "cure." Osteoarthritis (OA) is an example of a chronic pathology that results in fluctuating, often unpredictable, joint pain and functional limitations. (2,3) The fact that no disease-modifying therapies have been identified for OA may compel the person who is affected to more fully integrate coping strategies from the psychosocial and spiritual domains. In time, these domains may become inseparable components of the pathology itself, providing a complex relationship between the pathology and functional outcomes. (4,5) The purpose of my study was to examine the relative influences of physiological and psychosocial factors on functional outcomes in people with OA of the knee.

The Disablement Process Model (6) (Fig. 1) provides a framework for examining biomedical and psychosocial influences on functional outcomes in the presence of a pathology such as OA. Relationships have been established between the impairments of pain and functional limitations in walking (3,7) and between balance and functional limitations in walking. (8-11) In addition, psychosocial factors such as self-efficacy have been related to functional limitations in people with pain. (12-14) Brady et al (15) and Lorig and Holman16 provided additional insights into the importance of psychosocial factors such as self-efficacy in influencing outcomes for people with OA who participate in exercise programs and other self-management strategies.

[FIGURE 1 OMITTED]

Given the aging demographic, the problems created by OA for individuals or society will increase. Guccione (17) reported that the risk of disability attributable to OA of the knee alone was as large as that attributable to any other common medical condition such as heart disease or stroke in older adults. The National Health and Nutrition Examination Survey showed that OA of the knee was present in 10% to 20% of people 65 to 74 years of age, (3) and Felson et al (18) reported the occurrence to be 30% or greater in the population more than 75 years of age.

Pain is the primary stressor that motivates the person with arthritis to seek medical care. (19) Great strides in understanding the physiology of pain and its clinical management have been made in the past few decades. (5,20-22) Melzack (21) proposed that each person has a unique neuromatrix, or neurosignature, for the experience of pain that has integrated patterns of neural networks as its physiological basis. Although a person's response to pain (neuromatrix) is initially determined through genetics and early sensory development, life experiences related to pain and coping modulate and ultimately shape these neural patterns and activate certain perceptual and behavioral responses to pain that are unique to that person. (21) I believe that the concept of a neuromatrix provides a conceptual foundation for understanding the physiological integration of biomedical and psychosocial influences on the functional outcomes of people with chronic pain due to OA.

Social cognitive theory, as delineated by Bandura, (23) suggests that a person's beliefs about the world are primary determinants of his or her behavior. Self-efficacy is a construct used in social cognitive theory and is defined as a person's belief about his or her ability to successfully complete a task or activity. (23) According to social cognitive theory, self-efficacy influences the types of new challenges the person will try, how much effort will be expended, and how long the effort will persist in the face of obstacles. (24) Relationships have been established between self-efficacy and a variety of functional measures, including those of balance, instrumental activities of daily living, and exercise adherence. (15,24-27)

More research is needed to examine models in which psychosocial and biomedical factors are integrated to determine the relationship of pathology to functional limitations in older adults with OA. The aim of my analysis was to determine: (1) the relationships among OA, pain intensity, balance impairment, and the psychosocial factor of self-efficacy and (2) the relative effects of pathology, pain, balance, and self-efficacy on physical function in a sample of older women with OA of the knee.

Method

Design

The study was a post hoc analysis of subsets of data from a research study designed to examine the influence of pathology, pain, and coping on function in older women with OA of the knee. (28) A cross-sectional design was used with correlational analyses and regression modeling to determine the relative influences of pathology, pain, self-efficacy, and balance on physical function. Radiographs, symptom scales, the Functional Reach Test (FRT), (29) physical performance tests, and standardized questionnaires were used to quantify the independent variables of pathology, pain, balance, and self-efficacy and the dependent variables related to function.

Participants

A separate examination of men and women was necessary because sex differences have been demonstrated in the areas of pain expression, (30) OA occurrence, (31) and emotional responses to pain. (32) Because of the substantially increased occurrence of OA of the knee among women, my project focused on women. The frequency of OA begins to increase dramatically in the middle years, (3) and physical function (the dependent variable in my project) changes with age, (33,34) making age an important consideration. Inclusion criteria were: a community-dwelling female, 50 to 85 years of age, with symptoms of OA of the knee. The definition of symptoms of OA of the knee was consistent with the American College of Rheumatology's (ACR) clinical classification criteria. These criteria were developed by Altman and colleagues (35) through a detailed analysis and comparison of clinical, radiological, and laboratory characteristics of 264 patients with knee pain.

The final clinical classification criteria included knee pain plus at least 3 of the following: age over 50 years, stiffness, crepitus, bony tenderness, bony enlargement, and no palpable warmth. For the sample examined by Altman and colleagues, (35) these criteria had a 95% sensitivity and a 69% specificity in terms of determining OA of the knee. Age over 50 years was an inclusion criterion for my study; therefore, all volunteers for this study were initially screened using "knee pain" plus 2 additional characteristics from the ACR classification scheme. Volunteers with a medical diagnosis of OA of the knee (according to participant reports) also were included in the study if they met the other inclusion and exclusion criteria. Exclusion criteria were: more than 2 errors scored on the Short Portable Mental Status Questionnaire (SPMSQ) (36) (which was used to rule out people with cognitive impairments who would have had difficulty understanding the questionnaires), inability to walk independently (use of an assistive device, however, was permitted), and a medical history of stroke or other disorder that substantially affected lower-extremity function.

Participants were recruited through newspaper and radio advertising and through educational presentations at senior centers. Sixty-five people were screened via telephone interview for potential participation. Fifty-three participants came to the research center for data collection. Two participants dropped out of the study for unexplained reasons, and one died prior to completion of the study. Fifty community-dwelling women completed the study. Prior to data collection, a consent form approved by the University Medical Internal Review Board was explained to each participant and signatures were obtained.

The average age of the participants was 69.2 years (SD=8.8, range=52-84). The average weight was 75.1 kg (165.5 lb) (SD=16.3, range=39.9-113.4). The average height was 164.1 cm (64.6 in) (SD=6.1, range=150-176.3). The average educational level was "some college education," and the average annual income level was in the $25,000 to $35,000 range. Two subjects were unable to complete the radiological measures for the knees due to time issues in the radiology department, and 2 subjects were unable to complete the balance measure of functional reach due to their inability to get to the laboratory.

Measures

Cognitive status and demographics. The SPMSQ is a cognitive status test containing 10 questions, including date, day, location, telephone number, address, mother's maiden name, and a math procedure. Criteria have been established for the SPMSQ scores in identifying cognitive deficits. (36) Pfeiffer (36) examined intrarater reliability in 2 samples (n=30 and n=29) of adults over the age of 65 years. The Pearson statistic for test-retest correlation over a 4-week period was r=.82 and r=.83, respectively, for each sample, indicating sufficient stability of response for the purposes of this study. Data on participant characteristics (eg, age, ethnicity, daily activities, family support, use of prescribed and over-the-counter medications, education, income, past medical history) were collected via interview.

Pathology. Radiological films using a bilateral, standing anterior-to-posterior view of both knees were taken by a radiologist. Although the use of magnetic resonance imaging is being examined and its use in the diagnostic imaging for OA of the knee is growing, the standing anterior-to-posterior radiographic view continues to be the standard used for imaging the tibiofemoral joint. (31,37) Brandt (31) provided an overview of the difficulties of this approach to measurement, including inconsistency in patient and joint positioning, but conceded that this radiographic view has been considered the "gold standard" for examining OA of the knee clinically and in research for many years. Because it is commonly used by physicians in combination with a clinical exam in diagnosing OA of the knee and because of the assumption that medical management approaches to address functional limitations are based in part on medical diagnoses, it was used to grade pathology in my study.

At the end of data collection, a radiologist who was not given identifying information regarding the participants graded each knee film based on the Kellgren-Lawrence grading scale, a 0 to 4 scale for OA of the knee. (38) The grades for this scale are as follows: 0=no features of OA, 1=questionable osteophyte development, 2=definite osteophyte observed with minimal loss of joint space, 3=moderate loss of joint space width, and 4=severe loss of joint space width with subchondral bone sclerosis. (31,38) This scale is reported to be the most commonly used scale for grading OA of the knee and is recommended by the World Health Organization as the standard for cross-sectional and longitudinal studies. (31,37,39) Scott and colleagues (40) reported intraclass correlation coefficients (ICCs) between .82 and .95 for intrarater reliability and .63 and .83 for interrater reliability for 4 readers who used the Kellgren-Lawrence scale to grade 30 standing anterior-to-posterior knee radiographs of adults, aged 41 to 84 years, with radiological evidence of OA of the knee.

Pain and self-report of function. The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (41) is a 24-item scale that uses a 100-mm visual analog scale for participant scoring of each item. It is divided into 3 subscales and is designed to measure intensity of pain, stiffness, and functional difficulty in people with OA of the knee or hip. Lower scores indicate less pain, less functional difficulty, and less stiffness. The original instrument was developed with input from experts in rheumatology and patients, and an examination of currently existing measuring instruments also was made during its development. Factor analysis was used to create the subscale structure. (41) Construct validity is supported by the demonstration of its ability to discriminate between positive and negative outcomes in people with knee surgery, in people with OA of the hip or knee in terms of their levels of mobility, and in patients with mild or moderate OA versus those with severe OA. (42,43) The measures have been shown to be more specific to change in patients with OA than the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36). (44) Intraclass correlation coefficients for test-retest reliability over a 3-week interval with 60 adults aged 57 to 94 years diagnosed with OA of the knee were .95, .90, and .92 for the pain, stiffness, and function subscales, respectively. (45) The pain and function subscales were used in my study to quantify pain intensity and functional limitations.

Self-efficacy. The Arthritis Self-Efficacy Scale (ASES) (46) is a questionnaire consisting of 20 items characterizing the participant's belief that he or she can complete tasks related to pain management and physical function. The ASES consists of 3 subscales: self-efficacy in pain management, self-efficacy in function, and self-efficacy in other symptom management. The pain and function subscales were used in my study. The participants rated their "certainty" of their ability to perform each task on a 100-mm visual analog scale, with 0 being "very uncertain" and 100 being "very certain." Validity of data obtained with the instrument was supported in a study by Lorig and colleagues, (46) who described the development of the instrument, including focus group input from people participating in an arthritis self-management course (ASMC), input from content experts, quantitative question generation, factor analysis for subscale determination with elimination of unrelated questions, and analysis of internal consistency for the questions comprising each subscale. There is no "gold standard" for comparison with the ASES, although Lorig and colleagues (46) demonstrated in a sample of 144 people (average age=62.8 years, SD=13.2) participating in an ASMC that data for the self-efficacy in pain management subscale correlated with data for a 10-cm visual analog scale (r=.39) and that data for the self-efficacy in function subscale correlated with data for a measure of disability (r=.71). Using Pearson correlation coefficients (a less-than-ideal measure of reliability), researchers (46,47) demonstrated test-retest reliability in 91 people (average age=62.8 years, SD=13.2) participating in an ASMC to be .87 for pain management self-efficacy and .85 for functional self-efficacy.

Balance. The FRT (29) was used to screen for balance abilities. In this test, the person stands and reaches forward as far as possible, and the distance reached is measured. Three measurements were obtained in my study, and an average of the 3 measurements was taken as the final measurement. Validity of data for the FRT was demonstrated by a significant correlation (r=.71) with laboratory center of pressure measurements in a study of 128 people (aged 21-87 years) (29) and by the FRT's ability to discriminate fallers from nonfallers in a 6-month study of 217 men (aged 70-104 years). (48) In my study, within-session test-retest reliability was demonstrated by a Pearson correlation coefficient of .94.

Function: physical performance. Timed measurements of 3 functions were summed to provide a final physical performance score (in seconds) for each participant. The 3 timed functions were: (1) walking 20 m (use of a cane was permitted and timing began at a line 5 m from the subject's start position), (2) climbing up and down 9 stairs (use of handrail and either step-to or step-through patterns were permitted), and (3) going from sitting to standing for 5 repetitions (use of hands was not permitted, chair height was 43 cm [17 in], and timing ended when the subject's back touched the back of the chair on the last trial). These 3 measures were chosen because each represents an important function for older adults that can be affected by impairments such as knee pain, muscle force deficits, decreased range of motion, and decreased endurance, all of which are associated with OA of the knee. (17,49)

I believed that using all 3 measures was important in order to increase the complexity of the outcome measure and to add the challenge necessary for adequate variability (and to avoid a ceiling effect) among the community-dwelling older participants in my study. The literature supports the validity and reliability of data obtained for each as a functional measure with older community-dwelling adults. Self-selected walking speed was found to be a strong indicator of self-reported functional abilities. (50) Self-paced gait speed in 15 women (aged 37-74 years) with OA of the knee demonstrated intrarater reliability, with an ICC of .97 for the same session and an ICC of .88 with 1 week between sessions. (51) Timed sit-to-stand was shown to be associated with independence in instrumental activities of daily living in older adults, (52,53) to be able to discriminate between a group with rheumatoid arthritis and a group without arthritis, (54) and to be able to predict less disability 4 years later in older adults. (55) Timed sit-to-stand has been shown to be a reliable measurement method in a sample of 265 people aged 60 to 95 years, with an ICC of .71 for reliability with 2 to 3 weeks between sessions. (56) In a study of community-dwelling older adults, timed stair climbing had a correlation with balance and the Timed Up & Go Test, and a test-retest ICC value of .96 in older adults. (9)

Data Analysis

Descriptive information was examined through the use of means, standard deviations, frequencies, and ranges, as appropriate. I used the Pearson correlation coefficient to determine bivariate associations between continuous variables and the Spearman rank correlation for the average of the ranked variable of grade of OA of the knee. I used stepwise multiple regression modeling to determine the relationship of independent variables with the dependent variables of physical performance of function and self-report of functional difficulty, and I used analysis of variance calculations for beta weight testing. The acceptable level of significance for all relational statistical analyses was P [less than or equal to] .05.

Results

Pathology

Using the Kellgren-Lawrence grading scale (0-4), the mean scores for all participants were 2.04 (SD=1.28) for the right knee and 2.31 (SD=1.03) for the left knee. An average of the knee OA grades for the right and left sides was obtained for each individual, and the mean of these averaged scores was 2.16. Pathology grade was not correlated with data obtained for other variables (Table).

Pain Intensity

For all participants, the average pain measurement, using the WOMAC, was 39.2 (SD=21.3, range=6-87). Level of pain was negatively correlated with functional self-efficacy (r=-.48, P [less than or equal to] .001) and pain management self-efficacy (r=-.50, P [less than or equal to] .01) and was positively correlated with self-report of functional difficulty (r=.80, P [less than or equal to] .001) (Table).

Self-efficacy

The mean level of functional self-efficacy was 62.6 (SD=24.3, range=15-100) on the ASES. The mean level of pain management self-efficacy was 63.9 (SD=15.7, range=10-100). Pain management self-efficacy was correlated (in the expected directions) with level of pain, self-report of functional difficulty, and functional self-efficacy. Functional self-efficacy was correlated (in the expected directions) with pain level, self-report of functional difficulty, balance, and physical performance of function (Table).

Function

The mean physical performance measurement was 59.1 seconds (SD=26.7, range=23-155). The mean score for self-report of functional difficulty was 36.2 (SD=21.8, range=1.24-82.8). Higher scores on each measure were indicative of poorer performance. Self-report of functional difficulty and physical performance of function were correlated (r=.42, P [less than or equal to] .01). Self-report of functional difficulty also was correlated (in the expected directions) with pain management self-efficacy, functional self-efficacy, balance, and pain. Physical performance of function was correlated (in the expected directions) with functional reach, age, and functional self-efficacy (Table).

Stepwise Multiple Regression Analysis: Physical Performance of Function as the Dependent Variable The 4 independent variables with correlations with physical performance of function (age, balance, functional self-efficacy, and self-report of functional difficulty) were examined in regression modeling to determine their relative influence on physical performance of function. In the final model, functional self-efficacy and balance accounted for 42% of the variance in physical performance of function ([R.sup.2]=.42). The standardized beta estimates were [beta] =-.27 (P<.05) for relative influence of functional self-efficacy and [beta]=-.48 (P<.001) for balance, indicating that balance accounted for more of the variance in the model than self-efficacy did. As balance or functional self-efficacy increased, physical performance time decreased, or improved (Fig. 2).

[FIGURE 2 OMITTED]

Self-Report of Functional Difficulty: The Dependent Variable

The 5 independent variables with correlations with self-report of functional difficulty (ie, physical performance of function, pain, balance, functional self-efficacy, and pain management self-efficacy) were examined in regression modeling to determine their relative influence on self-report of functional difficulty. In the final model, pain and functional self-efficacy accounted for 74% of the variance ([R.sup.2]=.74, P<.0001). The standardized beta estimates were [beta]=.63 (P<.001) for relative influence of pain and [beta]=.34 (P<.001) for functional self-efficacy, indicating that pain accounted for more variance in the model for self-report of functional difficulty than did self-efficacy. As pain level increased or functional self-efficacy decreased, self-report of functional difficulty increased (Fig. 3).

[FIGURE 3 OMITTED]

Discussion

Sample

Overall, my sample comprised well-educated women with middle or higher annual incomes. Socioeconomic status is positively correlated with functional health and provides yet another source of varying influence on the functional outcomes of older adults with chronic pathology. (57) The relative homogeneity of the education and income levels of the sample may have minimized the influence of these factors on the variance in functional outcome.

Pathology

Pathology level was not predictive of function (self-report or physical performance). This finding may be due to the fact that the source of OA pain may not be the source of the pathology because articular cartilage is neither innervated nor vascularized. It is the reaction and interaction of other structures such as the joint capsule, the synovial lining of the capsule, the subchondral bone, and the muscle that result in nociceptor activation and, therefore, pain. (49) Although radiological examination of joint space width is commonly used by physicians to diagnose OA, some researchers question this technique in terms of its validity in predicting pain and function. (31) My study contributes to that line of questioning. Use of magnetic resonance imaging to determine OA pathology is growing as a valid and sensitive diagnostic approach, in part because of its ability to image cartilage and soft tissues. (58) Brandt (31) suggested that a more detailed clinical examination of the knee may result in an improved understanding of the tissue causing the pain and, thus, a more directed intervention approach.

Pain

The literature is replete with information associating chronic pain with negative patient status such as disability and depression. (5,59,60) Recognition of the negative impact of pain has grown to such an extent that the Department of Veterans Affairs has "declared war" on pain by making it the "fifth vital sign." (59) The sensory dimension of pain in my study was related to self-report of functional difficulty but not to physical performance of function. This finding suggests that self-report and physical performance of function represent separate domains that need to be measured individually. Pain level was correlated in the expected directions to pain management self-efficacy and functional self-efficacy, underscoring the importance of self-efficacy to the neuromatrix of pain.

Balance

A balance screening test was included among the independent variables because of its expected relationship to function. In the original study (from which these data were taken), balance was not a primary focus, and the FRT was considered an adequate screening test for balance. A more complex measure of dynamic balance may have yielded different results in this post hoc analysis, but it was not available. Nevertheless, balance as measured by the FRT was a variable in the final model for physical performance of function, but not for self-report of functional difficulty. Age dropped out of the regression analysis when placed with balance in the model for physical performance of function, suggesting that age is correlated with physical performance of function in part because of its relationship with balance.

Self-efficacy

Bandura (23) noted that self-efficacy is domain specific and a person may have strong self-efficacy in one domain and not in another, and this is supported by the low correlation between functional self-efficacy and pain management self-efficacy (r=.28, P [less than or equal to] .05). Functional self-efficacy accounted for both self-report and physical performance of function. The results of the regression analysis imply that, if increased function is desired, then increased functional self-efficacy also is desirable.

In other studies, (12,14,61,62) self-efficacy was reported to be a predictor of depression, pain level, and physical functioning in people with chronic pain. Rejeski and colleagues (14) demonstrated an interaction effect between knee extension force and functional self-efficacy in predicting self-reported disability in older adults with knee pain. Seeman and colleagues (62) concluded that functional self-efficacy has a large effect on people with weaker functional abilities, and they suggested that the difference seen between self-report of function and actual physical performance of function may be attributable to functional self-efficacy, with a greater difference suggesting lower self-efficacy. A modest, but significant, correlation existed in my study between physical performance and self-report of function (r=.42). Further analysis of my data revealed that participants with lower functional self-efficacy did not have a correlation between physical performance and self-report of function. Participants with higher functional self-efficacy, however, had a correlation between these 2 domains of function. This finding suggests that an accurate self-assessment of one's limitations, combined with adequate self-efficacy to address functional challenges, may be a desirable goal for people wanting to decrease pain and increase function.

The question remains: Can one improve self-efficacy, whatever the domain? Bandura (23) and others who subscribe to the tenets of cognitive behavioral theory (24) devised 5 key principles to use as a foundation for increasing self-efficacy: (1) increased opportunities for the person to experience successful performance of a task, (2) observation of role models who are successfully performing a task, (3) emotional arousal of the intention and desire to perform a task (ie, increased motivation), (4) a reduction in the negative physical feedback such as pain and fatigue associated with a task, and (5) verbal persuasion. The first principle, I believe, is part of functional retraining practiced by physical therapists. The second principle, observation of role models, is achieved through group approaches to education and rehabilitation. The third principle, emotional arousal to increase motivation, may best be achieved by supporting the person in prioritizing his or her own goals for therapy. Reducing the negative feedback of the experience may be achieved through physical therapy techniques directed toward pain management such as use of manual therapy for joint mobilization, (63) use of modalities (eg, heat, cold, and electrical stimulation), education regarding activity modification, use of assistive devices and proper shoe wear, and therapeutic exercise. 64 Verbal persuasion pertains not only to encouragement from voices of authority, but also to education of the person about the pathology and about what to expect from it. Demystifying the process may be helpful in returning control to the patient and increasing self-efficacy.

Limitations of Study

The primary limitation of my study was its cross-sectional design and the limited number of participants. Although regression modeling provided a stronger base for the existence of relationships between the independent and dependent variables, a longitudinal examination of the variables is needed to establish causality. In addition, although the ASES is considered the "gold standard" for validating arthritis-related self-efficacy instruments, the validity of data obtained with measurement approaches for a complex construct such as self-efficacy have been questioned. (65) Intervention studies also are needed to determine which recommendations for examination and intervention are effective. In addition, the use of a standing anterior-to-posterior radiograph to determine the grade of pathology, although commonly used clinically, provides information about the tibiofemoral joint, but not the patellofemoral joint. Future studies using advanced imaging techniques may provide more accurate information about the degree of pathology and its relationship to functional limitation. In addition, the use of a simple balance screening test limited the depth with which balance abilities could be explored.

Conclusion

The psychosocial factor of self-efficacy accounted for part of the variance for both domains of function in women with knee OA, while grade of pathology as measured in this study did not account for either domain. Balance accounted for part of the variance in physical performance of function, which was consistent with the information found in the literature review, while pain accounted for part of the variance in self-report of function. My research suggests that self-efficacy is likely to be influential in determining physical function. Rehabilitation, I believe, provides an important venue for addressing not only factors such as balance and pain, but also the psychosocial domain of self-efficacy as therapists support people with OA of the knee in making gains in function.

Table.
Correlations Among Pain, Function, Balance, and Self-efficacy
in Women With Osteoarthritis of the Knee (a)

                    Pathology   Age      Balance    Pain
Variable            (n=48)      (n=50)   (n=48)     (n=50)

Pathology           1            .17     -.23        .06
Age                             1         .52 (d)   -.12
Balance                                  1          -.26
Pain                                                1
Function
  Self-report
  Performance
Self-efficacy
  Functional
  Pain management

                    Function                    Self-efficacy

                                                             Pain
                    Self-report   Performance   Functional   Management
Variable            (n=50)        (n=50)        (n=50)       (n=50)

Pathology            .20           .27          -.26          .08
Age                 -.15           .36 (c)      -.21          .20
Balance             -.35 (b)      -.60 (d)       .48 (c)      .05
Pain                 .80 (d)       .22          -.48 (d)     -.50 (c)
Function
  Self-report       1              .42 (c)      -.66 (c)     -.35 (b)
  Performance                     1             -.48 (d)      .04
Self-efficacy
  Functional                                    1             .28 (b)
  Pain management                                            1

(a) Pearson correlation coefficient used throughout, except pathology
correlations, which use the Spearman rank correlation.

(b) P [less than or equal to] .05.

(c) P [less than or equal to] .01.

(d)P [less than or equal to] .001.

References

(1) Fried LP, Storer DJ, King DE, Lodder R. Diagnosis of illness presentation in the elderly. J Am Geriatr Soc. 1991;39:117-123.

(2) Affleck G, Tennen H, Keefe FJ, et al. Everyday life with osteoarthritis or rheumatoid arthritis: independent effects of disease and gender on daily pain, mood, and coping. Pain. 1999;83:601-609.

(3) Ettinger WH Jr, Afable RF. Physical disability from knee osteoarthritis: the role of exercise as an intervention. Med Sci Sports Exerc. 1994;26:1435-1440.

(4) Kleinman A. The Illness Narratives: Suffering, Healing. and the Human Condition. New York, NY: Basic Books; 1988.

(5) Turk DC. Biopsychosocial perspective on chronic pain. In: Gatchell RJ, Turk DC, eds. Psychological Approaches to Pain Management: A Practitioner's Handbook. New York, NY: Guilford Press; 1996:3-32.

(6) Verbrugge LM, Jette AM. The disablement process. Soc Sci Med. 1994;38:1-14.

(7) Davis MA, Ettinger WH, Neuhaus JM, Mallon KP. Knee osteoarthritis and physical functioning: evidence from the NHANES I Epidemiologic Follow-up Study. J Rheumatol. 1991;18:591-598.

(8) Rantanen T, Guralnik JM, Ferrucci L, et al. Coimpairments: strength and balance as predictors of severe walking disability. J Gerontol A Biol Sci Med Sci. 1999;54:M172-M176.

(9) Hughes C, Osman C, Woods AK. Relationship among performance on stair ambulation, functional reach, and timed up and go. Issues on Aging. 1998;21(3):18-22.

(10) Duncan PW, Chandler J, Studenski S, et al. How do physiological components of balance affect mobility in elderly men? Arch Phys Med Rehabil. 1993;74:1343-1349.

(11) Jette AM, Assman SF, Rooks D, et al. Interrelationships among disablement concepts. J Gerontol A Biol Sci Med Sci. 1998;53: M395-M404.

(12) Arnstein P, Caudill M, Mandle CL, et al. Self efficacy as a mediator of the relationship between pain intensity, disability and depression in chronic pain patients. Pain. 1999;80:483-491.

(13) Barlow JH, Williams B, Wright C. The Generalized Self-Efficacy Scale in people with arthritis. Arthritis Care Res. 1996;9:189-196.

(14) Rejeski WJ, Miller ME, Foy C, et al. Self-efficacy and the progression of functional limitations and self-reported disability in older adults with knee pain. J Gerontol B Psychol Sci Soc Sci. 2001;56:S261-S265.

(15) Brady TJ, Kruger J, Helmick CG, et al. Intervention programs for arthritis and other rheumatic diseases. Health Educ Behav. 2003;30: 44-63.

(16) Lorig KR, Holman H. Self-management education: history, definition, outcomes, mechanisms. Ann Behav Med. 2003;26:1-7.

(17) Guccione AA. Osteoarthritis, comorbidity, and physical disability. In: Hamerman D, ed. Osteoarthritis: Public Health Implications for an Aging Population. Baltimore, Md: Johns Hopkins University Press; 1997:84-98.

(18) Felson DT, Lawrence RC, Dieppe PA. Osteoarthritis: new insights, part 1: the disease and its risk factors. Ann Intern Med. 2000;133: 635-646.

(19) Mahat G. Perceived stressors and coping strategies among individuals with rheumatoid arthritis. J Adv Nurs. 1997;25:1144-1150.

(20) Craig KD. Emotions and psychobiology. In: Wall PD, Melzack R, eds. Textbook of Pain. 4th ed. New York, NY: Churchill Livingstone; 1999:331-343.

(21) Melzack R. From the gate to the neuromatrix. Pain. 1999;6(suppl 6):S121-S126.

(22) Mostofsky DI, Lomranz J. The psychology of pain and suffering. In: Mostofsky DI, Lomranz J, eds. Handbook of Pain and Aging . New York, NY: Plenum Press; 1997:3-18.

(23) Bandura A. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev. 1977;84:191-215. (24) Cousins SO. Exercise, Aging, and Health. Washington, DC: Taylor & Francis; 1998.

(25) Maki BE, Holliday PJ, Topper AK. Fear of falling and postural performance in the elderly. J Gerontol. 1991;46:M123-M131.

(26) McAuley E, Lox C, Duncan TE. Long-term maintenance of exercise, self-efficacy, and physiological change in older adults. J Gerontol. 1993;48:P218-P224.

(27) Myers AM, Powell LE, Maki BE, et al. Psychological indicators of balance confidence: relationship to actual and perceived abilities. J Gerontol A Biol Sci Med Sci. 1996;51:M37-M43.

(28) Harrison AL. The Influence of Pathology, Pain and Coping on Function in Women With Knee Osteoarthritis [dissertation]. Lexington, Ky: University of Kentucky, Program in Gerontology; 2002.

(29) Duncan PW, Weiner DK, Chandler J, Studenski S. Functional reach: a new clinical measure of balance. J Gerontol. 1990;45:M192-M197.

(30) Unruh AM. Gender variations in clinical pain experience. Pain. 1996;65:123-167.

(31) Brandt KD. Diagnosis and Nonsurgical Management of Osteoarthritis. Cado, Okla: Professional Communications Inc; 2000.

(32) Keefe FJ, Lefebvre JC, Egert JR, et al. The relationship of gender to pain, pain behavior, and disability in osteoarthritis patients: the role of catastrophizing. Pain. 2000;87:325-334.

(33) Buchner DM, Cress ME, Esselman PC, et al. Factors associated with changes in gait speed in older adults. J Gerontol A Biol Sci Med Sci. 1996;51:M297-M302.

(34) Winter DA, Patla AE, Frank JS, Walt SE. Biomechanical walking pattern changes in the fit and healthy elderly. Phys Ther. 1990;70: 340-347.

(35) Altman RD, Asch E, Bloch D, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association. Arthritis Rheum. 1986;29: 1039-1049.

(36) Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc. 1975;23:433-441.

(37) Link TM, Steinbach LS, Ghosh S, et al. Osteoarthritis: MR imaging findings in different stages of disease and correlation with clinical findings. Radiology. 2003;226:373-381.

(38) Kellgren JH, Lawrence JS. Radiologic assessment of osteo-arthrosis. Ann Rheum Dis. 1957;16:494-502.

(39) Altman RD. Criteria for classification of clinical osteoarthritis. J Rheumatol Suppl. 1991;27:10-13.

(40) Scott WW Jr, Lethbridge-Cejku M, Reichle R, et al. Reliability of grading scales for individual radiographic features of osteoarthritis of the knee: the Baltimore Longitudinal Study of Aging atlas of knee osteoarthritis. Invest Radiol. 1993;28:497-501.

(41) Bellamy N, Buchanan WW, Goldsmith CH, et al. Validation study of the WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol. 1988;15: 1833-1840.

(42) Bombardier C, Melfi CA, Paul J, et al. Comparison of a generic and a disease-specific measure of pain and physical function after knee replacement surgery. Med Care. 1995;33(4 suppl):AS131-AS144.

(43) Ehrich EW, Schnitzer TJ, McIlwain H, et al. Effect of specific COX-2 inhibition in osteoarthritis of the knee: a 6 week double blind, placebo controlled pilot study of rofecoxib. Rofecoxib Osteoarthritis Pilot Study Group. J Rheumatol. 1999;26:2438-2447.

(44) Soderman P, Malchau H. Validity and reliability of Swedish WOMAC osteoarthritis index: a self-administered disease-specific questionnaire (WOMAC) versus generic instruments (SF-36 and NHP). Acta Orthop Scand. 2000;71:39-46.

(45) Dunbar MJ, Robertsson O, Ryd L, Lidgren L. Appropriate questionnaires for knee arthroplasty: results of a survey of 3600 patients from the Swedish Knee Arthroplasty Registry. J Bone Joint Surg Br. 2001;83: 339-344.

(46) Lorig K, Chastain RL, Ung E, et al. Development and evaluation of a scale to measure perceived self-efficacy in people with arthritis. Arthritis Rheum. 1989;32:37-44.

(47) Parker JC, Smarr KL, Buckelew SP, et al. Effects of stress management on clinical outcomes in rheumatoid arthritis. Arthritis Rheum. 1995;38:1807-1818.

(48) Duncan P, Studenski S, Chandler J, Prescott B. Functional reach: predictive validity in a sample of elderly male veterans. J Gerontol. 1992;47:M93-M98.

(49) Threlkeld AJ, Currier DP. Osteoarthritis: effects on synovial joint tissues. Phys Ther. 1988;68:364-370.

(50) Cress ME, Schechtman KB, Mulrow CD, et al. Relationship between physical performance and self-perceived physical function. J Am Geriatr Soc. 1995;43:93-101.

(51) Marks R. Reliability and validity of self-paced walking time measures for knee osteoarthritis. Arthritis Care Res. 1994;7:50-53.

(52) Judge JO, Schechtman K, Cress E. The relationship between physical performance measures and independence in instrumental activities of daily living. J Am Geriatr Soc. 1996;44:1332-1341.

(53) Gill TM, Williams CS, Tinetti ME. Assessing risk for the onset of functional dependence among older adults: the role of physical performance. J Am Geriatr Soc. 1995;43:603-609.

(54) Jette AM, Jette DU, Ng J, et al. Are performance based measures sufficiently reliable for use in multicenter trials? The Musculoskeletal Impairment Study (MSI) Group. J Gerontol A Biol Sci Med Sci. 1999;54: M3-M6.

(55) Guralnik JM, Ferrucci L, Simonsick EM, et al. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332:556-561.

(56) Ostchega Y, Harris TB, Hirsch R, et al. Reliability and prevalence of physical performance examination assessing mobility and balance in older persons in the US: data from the Third National Health and Nutrition Examination Survey. J Am Geriatr Soc. 2000;48:1136-1141.

(57) George LK. Social factors and illness. In: Binstock RH, George LK, eds. Handbook of Aging and the Social Sciences. San Diego, Calif: Academic Press; 1996:229-254.

(58) Andriacchi TP, Lang PL, Alexander EJ, Hurwitz DE. Methods for evaluating the progression of osteoarthritis. J Rehabil Res Dev. 2000;37: 163-170.

(59) Flaherty JH. "Who's taking your fifth vital sign?" J Gerontol A Biol Sci Med Sci. 2001;56:M397-M399.

(60) Turk DC, Okifuji A, Scharff L. Chronic pain and depression: role of perceived impact and perceived control in different age cohorts. Pain. 1995;61:93-101.

(61) Asghari A, Nicholas MK. Pain self-efficacy beliefs and pain behaviour: a prospective study. Pain. 2001;94:85-100.

(62) Seeman TE, Unger JB, McAvay G, Mendes de Leon CF. Self-efficacy beliefs and perceived declines in functional ability: MacArthur Studies of Successful Aging. J Gerontol B Psychol Sci Soc Sci. 1999;54:P214-P224.

(63) Deyle GD, Henderson NE, Matekel RL, et al. Effectiveness of manual therapy and exercise in osteoarthritis of the knee: a randomized, controlled trial. Ann Intern Med. 2000;132:173-181.

(64) Guide to Physical Therapist Practice. 2nd ed. Phys Ther. 2001;81: 9-746.

(65) Brady TJ. Measures of self-efficacy, helplessness, mastery, and control. Arthritis Rheum. 2003;15:S147-S164.

AL Harrison, PT, PhD, is Associate Professor, Department of Rehabilitation Sciences, Division of Physical Therapy, College of Health Sciences, University of Kentucky, Lexington, KY 40506 (USA) ([e-mail]).

This research is a post hoc analysis of a subset of data from Dr Harrison's dissertation research for her Doctoral Degree in Gerontology at the University of Kentucky.

The author thanks Nancy Stiles, MD; Julia Luan, MS, biostatistician; Benedek Bognar, MD, radiologist; and Agnes Bognar, BS, RT, R(MR).

This work was presented as a poster presentation at the Combined Sections Meeting of the American Physical Therapy Association; February 4-8, 2004; Nashville, Tenn.

This project was funded internally by the University of Kentucky College of Health Sciences and was approved by the University of Kentucky Office of Research Integrity and Medical Internal Review Board.

This article was received November 18, 2003, and was accepted March 15, 2004.