During the early years of joint replacement, relief of the crippling pain of arthritic disease was the primary, and, to a large extent, the only requirement of these procedures. However, during the past 15 years, increasing emphasis has been given to the outcome of medical treatment, generally measured in terms of its real and perceived success in achieving pretreatment goals. Studies have shown that patients’ expectations of orthopaedic treatment are highly variable, and depend on the age, diagnosis, and lifestyle of each patient.8,11 These studies also have confirmed that these preoperative expectations strongly influence patient outcomes after total joint replacements. 8,11 To a considerable extent, the goals of individual patients influence whether the knee replacement operation is perceived as being successful, whether the patient thinks that they have significant residual disability, and whether symptoms related to knee function will cause the patient to seek additional treatment in the future. 3,6,7
In anticipating a successful outcome from treatment of any chronic degenerative disease, it is natural for the patient to hope that his or her function might be restored to what it was before the onset of the initial symptoms. This hope often supercedes many considerations, including the passage of time, the effects of age, and the extent to which the patient has maintained his or her health and fitness despite the disabling effects of their chronic disease. In the face of these hopes, the surgeon must guide the patient in formulating reasonable expectations for the outcome of each available treatment option. However, this is only possible if the surgeon knows the average outcome for all patients and the impact of any factors that may modify the outcome expected for a particular patient.
Previous attempts to quantify patient function after total knee replacement (TKA) have been limited by two main factors: most instruments assess only gait and stair climbing without consideration of the functional activities of importance to each patient; and no allowance is made for normal deterioration of musculoskeletal function that occurs with aging, despite the advanced age of many patients having TKA. This study addresses the following question: What are the effects of aging on knee function? The answer to this question will establish a realistic normal, healthy knee function benchmark for patients after TKA, allowing us to address the next question: What is the true ability of a knee arthroplasty in restoring normal, healthy knee function?
MATERIALS AND METHODS
In a previous study, a validated Total Knee Function Questionnaire was developed to assess the frequency with which patients performed a wide spectrum of activities involving the knee.13 This survey instrument consisted of 55 multiple choice questions regarding symptoms and physical activities. Physical activities were divided into three categories: (1) baseline activities (17 questions), defined as fundamental activities of daily living that required knee function, including walking, stair climbing, sitting, foot care, bathing, and car travel; (2) advanced activities (nine questions), which required greater strength, control, and ROM, including kneeling, squatting, moving laterally, turning and cutting, and carrying loads (such as a bag of groceries); and (3) recreational activities and exercises (16 questions), which consisted of 12 activities (stretching, stationary biking, leg strengthening, swimming, golfing, tennis, cross-country skiing, downhill skiing, dancing, gardening, sexual activity, and running). An additional 13 questions evaluated patient satisfaction, activity level, pain, and other symptoms.
In considering each activity, patients were asked to assess the personal importance of the activity and the extent to which their participation was hindered by their knee disorder. The patients’ ability to walk was explored in detail. The patients were asked how long they could walk without stopping, whether they used walking aids, about walking up and downhill, and about walking on even and uneven ground. Basic demographic information, overall health status, history of arthritis, and significant knee history also were requested in the questionnaire.
A modified version of the Total Knee Function Questionnaire was prepared through deletion of all items directly referring to knee arthroplasty. With approval from the Institutional Review Board at our institution, the modified instrument was distributed to volunteers (referred to as “subjects”) at various venues, including meetings of the Association for the Advancement of Retired Persons (AARP), outpatient general medicine clinics at several hospitals, and employees of our department and their relatives. To participate in the study, respondents had to be older than 40 years and had to be able to complete the questionnaire in English. Respondents were excluded from the study group if they had previous total knee surgery, if they took more than five prescription medications at the time of the questionnaire, or if they had any serious medical problems. This list included, but was not limited to, congestive heart failure, emphysema, severe asthma, or general central nervous system dysfunction. Participants received no remuneration for completion of the questionnaire.
The functional abilities of the volunteer group were compared with a similar group of patients who had total knee arthroplasties who participated in an earlier study using the Total Knee Function Questionnaire.13 This cohort consisted of 176 patients (referred to as “patients”) with 105 (60%) women and 71 (40%) men, who had TKAs at least 1 year previously. The average age of the patients was 71 ± 11 years for the women, and 70 ± 9 years for the men. Twenty-four percent of these patients were younger than 65 years, 41% were between 65 and 75 years, and 35% were older than 75 years.
For each recreational activity, a composite function score was developed by combining the frequency of each patient’s participation in that activity, its perceived importance to the patient, and the amount of hindrance, or bother, imposed by the TKA in doing that activity.13 The composite score ranges from −5–5. A negative bother score is given if the patient experiences any more than a slight amount of bother when performing that activity. A positive bother score indicates little or no bother with the highest score awarded for no bother. This number (−2–2) then is scaled by how important the activity is (0–5) and by how frequently it is done (0–5).
The significance of differences in activity prevalence, and subscale scores by gender, age, and walking ability, were assessed using an analysis of variance (ANOVA) test (when more than two groups were compared) and an unpaired t test (when only two groups were compared). Ordinal data describing respondents’ rankings of frequency, importance, and difficulty of individual activities (five-level scale) were compared between cohorts, and subgroups were stratified by age and gender. The statistical analysis of these comparisons was evaluated using the chi square test. In all comparisons, significant differences were defined as p < 0.05.
As a check of the validity of the responses obtained from the control subjects, the data derived from the questionnaire relating to the prevalence of participation in different activities were compared with the results of a much larger sample published from the Third National Health and Nutrition Examination Survey (NHANES III).4 This survey included responses from more than 19,000 men and women older than 50 years. Although the questions from the NHANES III survey were not identical to those in our Knee Function Questionnaire, similar questions were grouped regarding specific activities and symptoms. This allowed quality control of our data to avoid skewing by sampling nonrepresentative groups. In one study that used NHANES III, 4 the prevalence of knee pain was assessed among adults in the US older than 60 years.1
Two hundred fifty-seven completed questionnaires were received from control subjects. Forty-six percent of the respondents (117) were males, and 54% (140) were females (Table 1). The average age of the male respondents was 67 ± 11 years (range, 34–95 years), compared with 70 ± 11 years for the female respondents (range, 46–94 years). The respondents were recruited from Veteran’s Affairs clinics (n = 25), 17 orthopaedic department affiliations (n = 17), public hospital clinics (n = 50), private hospital clinics (n = 67), and AARP meetings throughout Houston, TX (n = 98).
Approximately 2/3 of the control subjects reported that they did stretching (63%) and leg-strengthening exercises (66%) at least occasionally, and performed activities involving moving laterally (69%), cutting and turning (70%), kneeling (63%), and squatting (59%; Fig 1). Other common activities were carrying heavy objects (46%), gardening (51%), engaging in sexual activity (41%), and dancing (27%). Nineteen percent of control subjects swam, at least occasionally, and 17% rode an exercise bike, although only 8% played golf. A small subset of less than 5% of control subjects participated occasionally in high-demand activities, including tennis and skiing. Surprisingly, 175 (68%) control subjects reported that they had run at least a few steps, and 12 (5%) of these subjects had run more than 1 mile.
Expressed as a percentage of all individuals who did each activity, the most important activities to control subjects were leg-strengthening exercises (76%), sexual activity (67%), golf (65%), kneeling (61%), stretching exercises (57%), and gardening (57%; Figure 2). Control respondents also reported that they considered their ability to turn or cut (61%), move laterally (56%), squat (52%), and carry heavy objects (60%) important, in addition to recreational activities including golf (65%), dancing (58%), and swimming (42%).
Many of the control subjects in this study reported that their knees performed completely normally, and that they could do most of the activities cited without symptoms attributable to their knees (Fig 3). Overall, almost all the control participants described their knee function as completely unimpaired when doing the following activities: swimming (88% of participants), sexual activity (86% of participants), and dancing (86% of participants). Almost as many of our control subjects reported that they could do stretching exercises (76% of participants), stationary biking (70% of participants), leg-strengthening exercises (71% of participants), and carry heavy objects (71% participants) without discomfort. Two-thirds (67%) of participants could garden without difficulty, however, approximately ½ could not kneel (54%) or squat (52%) without some knee pain. Of the small subset of individuals who participated in sporting activities, almost none experienced knee discomfort during golf (7% of participants), racquet sports (11% of participants), cross-country skiing (0% of participants), or swimming (5% of participants).
Relatively few control subjects reported moderate to severe difficulties with knee function when doing low-demand activities, including swimming (2% of participants), stationary biking (5% of participants), leg strengthening (3% of participants), sexual activity (5% of participants), dancing (1% of participants), gardening (5% of participants), or moving laterally (4% of participants; Figure 4). However, knee symptoms were experienced more frequently during activities that placed greater loads on the extremity, such as carrying heavy shopping bags (6% of participants), or those that demanded increased ROM, such as kneeling (12% of participants), squatting (11% of participants), or stretching exercises (11% of participants).
Overall, there was no difference (p > 0.05) in the knee function scores of men and women, however, male control subjects in both groups showed a general, continuous deterioration in knee function with increasing age (Fig 5). Detailed examination of the responses of the male respondents showed that the activities that caused the most difficulty were kneeling and squatting in men younger than 85 years, and sexual activity in men older than 85 years. More than 88% of men in all age groups were able to walk at least 15 minutes, with fewer than 5% of all control respondents using a walking aid (cane, walker, or wheelchair). More than 83% of male respondents in all age groups also were able to run a short distance. In women, the activities most limited by knee function were kneeling and squatting, independent of age. Almost all women younger than 75 years, and 85% of women between 75 and 85 years, could walk for more than 15 minutes compared with approximately ½ (55%) of the female participants older than 85 years. None of the female control respondents younger than 75 years used a walking aid, whereas aids were used for walking by 5% of women 75 years and older. Greater than 92% of women younger than 75 years reported being able to run a short distance, whereas 75% of women between 75 and 85 years, and only 27% of women older than 85 years, could run a short distance.
Our data compared well with the data from the NHANES III survey (Table 2).4 Specifically, responses were similar for men and women regarding the ability to walk 2–3 blocks without any difficulty. Also compared was the ability (or lack of limitation) of getting up from a low seat, in which the participants in the current study and the participants in the NHANES III survey had similar results. In both groups, respondents had increasing difficulty with stooping, crouching, and kneeling as age increased, and women reported more difficulty than men. The NHANES III group and the participants surveyed with the Total Knee Function Questionnaire in the current study reported minimal participation in more vigorous activities, including riding an exercise bicycle, swimming, aerobics, and aerobic dancing, with decreasing percentages as the respondents aged. Finally, similar percentages of respondents in both groups reported participation in yard work and gardening, which is a common activity for individuals as they age. We found that the results of control subjects in our study were consistent with results of the larger sample from the NHANES III survey. Namely, for questions in comparable areas of knee function, our new instrument, the Total Knee Function Questionnaire, correlated well with previously collected data on subjects with no knee disorders.
There were large differences between the group of patients who had TKA and the age- and gender-matched control subjects described above who had no previous knee disorders in their functional capacity to do activities involving the knee. This is best shown by the Composite Knee Function Scores, which averaged 5.7 ± 0.2 for the group of patients who had TKA, compared with 6.9 ± 0.3 for the control subjects across all activities (p < 0.00001). In comparative terms, two main groups of activities were identified: activities in which the patients and control subjects had essentially similar knee function (absolute difference in function score < 0.5; swimming, golfing, and stationary biking), and activities in which the function scores of the control group exceeded the scores of the patients who had TKA (kneeling, squatting, moving laterally, turning and cutting, carrying loads, stretching, leg strengthening, tennis, dancing, gardening, and sexual activity; Figure 6).
Overall, approximately ⅓ more control subjects with no previous knee disorders kneeled, compared with patients who had TKA (63% versus 42%, respectively; chi square = 13.0; p = 0.0003), or squatted (59% versus 35%, respectively; chi square = 17.0; p < 0.00001), whereas twice as many patients who had TKA participated in stationary biking compared with control subjects (33% versus 17%, respectively; chi square = 28.2; p < 0.00001; Figure 7). In general, activities that placed few demands on the knee, in terms of muscle control or ROM, were equally important to both groups and were cited by approximately ½ of those who were regular participants in these activities (Fig 8). These activities included gardening, stretching, swimming, golf, and sexual activity. Two to three times as many patients who had a knee replacements experienced difficulties doing most other activities attributable to impaired knee function (Fig 9). The patients who had TKAs were almost comparable to the control subjects when reporting symptom-free performance of stationary biking (53% versus 70%, respectively), sexual activity (64% versus 86%, respectively), and swimming (65% versus 88%, respectively). However, as the activities became more demanding, the gap between the two populations widened, with approximately four times as many control subjects reporting symptom-free function compared with patients who had TKAs when squatting (52% versus 14%, respectively) or kneeling (54% versus 13%, respectively; Fig 10).
Some surprising differences were seen when the average Composite Knee Function scores across all activities done by the control subjects and the patients who had TKAs were compared by age and gender (Fig 11). In men, the average Composite Knee Function score in both groups declined slightly with advancing age, although the average Composite Knee Function score in the patients who had TKAs was 10–20% less than that of the control subjects in each of the age ranges studied. Conversely, the average Composite Knee Function scores of the female patients who had TKAs increased dramatically in patients older than 85 years and was less (p < 0.01) in the youngest age group (younger than 55 years) when compared with the control subjects. This may be indicative of systematic bias in patient selection in these two age ranges, or may indicate that, in the patients studied, TKA failed to meet the expectations of the younger female patients but was successful in meeting the expectations of the older female patients.
Determining healthy, normal knee function in people 50 years and older is complicated. By describing our respondents’ level of participation in a particular activity, ascribed importance, and limitations for each activity, we have taken a first step toward determining healthy, normal knee function. In questioning patients regarding a diverse range of activities, we have obtained information regarding not only what the respondents are able and unable to do, but also of what is important to them. We think this is a critical step in assessing patient satisfaction for outcome studies. In this study, we established a baseline against which reports of knee function can be compared in the future. Moreover, the data we collected provide insight into how the knees of aging individuals function during a broad range of activities, from simple activities of daily living to high-demand activities such as racquet sports, skiing, and dancing. In these appreciable respects, the questionnaire used in the current study differs from other published outcome instruments used in previous studies. 2,5,8,10–12
Our study indicates that men and women older than 50 years do participate in various activities, and have much in common. Among men, the most common activities varied with age. Stretching exercises were done regularly (at least once per week) by men 55–65 years (67%), 65–75 years (68%), and older than 85 years (71%), but were done less frequently by male participants younger than 55 years (41%) and 75–85 years (47%). In women, regular stretching (at least once per week) was the most common activity in all age groups, ranging from 54% of female respondents 55–65 years old, to 81% of women younger than 55 years.
Our data show that control subjects do have problems with their knees, perhaps even more so than previously recognized. In men and women, knee function declined gradually with age, decreasing more rapidly in subject older than 85 years. In women, the ability to walk was a good indicator of the change in knee function. Although all women younger than 55 years in our study were able to walk more than 15 minutes, only 55% of women 85 years old and older were able to do so. The effect of aging in the male population was less dramatic in terms of aggregate Composite Knee Function scores, however, men reported greater dissatisfaction with their ability to do activities that were of greatest personal importance, especially when they reached 65 years of age.
In this study, the control subjects who had no previous knee disorders reported difficulties with their knees in many of the same activities as patients who had TKAs. Activities which traditionally have been perceived to be limited after TKA, such as kneeling, squatting, climbing stairs, and getting up from a low seat, were the most commonly limited activities among control subjects. This suggests that more similarities exist between the function of replaced knees and nonreplaced knees despite the objective differences (such as ROM and strength) than was previously thought.
Some of the most important conclusions of this study are evident when the responses of the control subjects with healthy knees are compared with those of the patients who had TKAs. Comparison of the two groups is possible because both groups of patients completed the same standardized questionnaire and were similarly distributed by age and gender. Few differences were seen between the two populations in terms of average frequency of participation in different activities, with the exception of the most demanding activities and stationary biking. A similar conclusion was shown by data describing the importance of each activity to the respondents in both groups. In general, activities that placed few demands on the knee, in terms of either muscle control or ROM, were equally important to both groups, and were cited by approximately ½ the control subjects and patients who were regular participants in these activities. Conversely, other activities, including squatting, dancing, carrying heavy objects, kneeling, and turning or cutting, were considered important by 20–40% more control subjects than patients who had TKAs. We suggest that this is because patients attenuate their activity and shift their attention to less demanding pursuits once they discover the biomechanical limitations of their replaced joint in terms of strength and stability.
Differences were observed between the control subjects and patients who had TKAs in the prevalence of individuals experiencing substantial difficulty in doing some of the functional activities in the survey. Two to three times as many patients who had a TKA reported some degree of difficulty in doing most of the activities in the survey attributable to limitations imposed by knee function. Similarly, only ½ as many respondents considered their knee function normal after knee replacement compared with age-matched controls. The patients who had TKAs and control subjects were only comparable when doing stationary biking (53% versus 70%, respectively), sexual activity (64% versus 86%, respectively), and swimming (65% versus 88%, respectively). As the activities became more demanding, the gap between the two populations widened, with approximately four times as many control subjects reporting symptom-free function when squatting (52% versus 14%) or kneeling (54% versus 13%) compared with patients who had TKAs.
On the basis of these data, it is clear that much of the impairment of function reported by patients with TKAs cannot be attributed to the effects of age or unrealistic expectations. In reality, the peers of patients who have had joint replacement experience less impairment in doing many activities involving the knee. Moreover, this difference becomes greater as the activities become more demanding. The underlying causes of these differences are unclear and are the subjects of additional investigation. Although it is likely that much of these differences is a reflection of the biomechanical deficiencies of contemporary knee replacement designs, other likely factors include alteration of the remaining soft tissues and the absence of the native cruciate ligaments. Related factors include the general condition of the soft tissues in the patients with TKAs, including the presence of scar tissue, changes attributable to OA, and a possible reduction in muscular tone and lower limb strength. Each of these factors would be expected to substantially compromise the functional capacity of the knee, especially when doing high-demand activities requiring strength, flexibility, or control.
An additional question is whether middle-aged or elderly individuals without indications of knee disorders should be regarded as healthy, therefore setting a standard for comparison or aspiration for patients who have degenerative joint disease. It is possible that the best prosthetic replacement done with minimal tissue disruption in the patient who is most motivated and rehabilitated will not provide function similar to that of individuals who have no degenerative changes. The function of the asymptomatic healthy joint may be a goal for reference purposes or for aspiration of patients with degenerative joint disease.
The current study has many limitations, which are well-recognized by all who do questionnaire-based research. First, self-administered instruments, by their nature, are inherently subjective, and so have limited ability to accurately assess knee function. Nonetheless, they are the simplest, least invasive, and most cost-effective way to obtain normative data.9 Second, the problem of responder bias must be considered in any survey-based research. To avoid skewing the responses obtained in the current study, public and private settings were selected, as were hospital and nonhospital settings, for the recruitment of our control group. Although it is impossible to know what the responses would have been from people who chose not to participate, we have no reason to think that there is any substantial responder or nonresponder bias. Furthermore, our comparison of results with the NHANES III database shows that our sample is not skewed in any direction. Third, the number of participants in the current study was limited to 257 respondents and in some categories we have less than 15 responses. A larger sample would yield greater statistical power and would allow us to make more definitive conclusions regarding the functional activities of specific subgroups, rather than overall trends.
In comparison with other outcome measures, the Total Knee Function Questionnaire differs philosophically from the Short Form-36 in that it deliberately focuses on tasks that place demands on the knee. The Total Knee Function Questionnaire does not attempt to evaluate a patient’s overall quality of life, but rather the dimensions of knee function that affect lifestyle, weighted by the importance of each dimension to the patient. In addition, the Total Knee Function Questionnaire does not assume, a priori, that the activities and priorities prescribed by the designer of the questionnaire are reflective of the priorities and lifestyle of each respondent. In practice, there was a weak correlation between patients’ functional outcome and the Short Form-36 and Knee Society scores. The Short Form-36 score is distorted by comorbidities and generalized inquiries concerning function having little to do directly with the knee. The Knee Society score is completed by the surgeon, and is based only on walking and stair-climbing. We have found that this results in a pronounced ceiling effect arising from patients who are given high scores on the Knee Society score but who still are limited significantly in doing many activities that they consider important.
The current study was done to provide objective information regarding the functional ability of control subjects and patients who have had TKAs to assist patients and surgeons in developing realistic expectations of the outcome of TKA. Using this information, we found that a TKA does not restore normal knee function, independent of the effects of age and gender. Although this procedure restores the patient’s ability to do many routine activities, a substantial deficit remains in meeting the challenges of many functional tasks that are important to the patient, especially tasks involving kneeling or squatting. Currently, we must caution patients that their knee will be pain-free after TKA, but will likely not function as well as before the onset of their arthritic symptoms. Our future success in improving knee function after joint replacement requires careful attention to many facets of the procedure, including the design and placement of the prosthetic components, the treatment and preservation of soft tissue structures, and the invasiveness of the surgical approach. However, some fundamental limitations always will exist because of the irreversible effects of disease and factors predisposing some individuals to joint degeneration. we think correction of these factors will be elusive for the foreseeable future.