Abstract
Purpose
To evaluate the effects of total knee arthroplasty (TKA) on symptoms, function and activity over 5 years in knee osteoarthritis (KOA) patients.
Methods
Data were from the Osteoarthritis Initiative (OAI). Participants who conducted the first TKA from (not before) enrolment to 48 months were propensity score matching (PSM) on their characteristics at the visit before surgery (treated as baseline) to those who did not conduct a TKA at 48 months (treated as baseline). Changes in knee pain and functional disability were evaluated using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain and function subscales. Changes in physical activity, physical function and overall physical health and mental health were assessed using the Physical Activity Scale for the Elderly, the 20‐m walk speed and the five times chair‐to‐stand tests and the 12‐Item Short Form Survey, respectively.
Results
Eighty‐two pairs of participants in the TKA (56% women, mean 64.8 ± 8.4 years) and non‐TKA groups were matched. Knee symptoms were significantly improved in the TKA group and flatted from 24 months (Pain: β = −3.29, 95% confidence interval [CI] = [−4.59 to −1.99], p < 0.001; Function: β = −10.12, 95% CI = [−14.21 to −6.03], p < 0.001). Physical function and overall physical health but not physical activity or mental health (PASE: β = 5.72, 95% CI = [−15.46 to 26.90], p = 0.597; Mental: β = 0.04, 95% CI = [−2.47 to 2.54], p = 0.976) was improved in the TKA group over 24 months.
Conclusions
TKA substantially improved knee symptoms and physical function over 60 months and physical health over 48 months, compared to those who had a similar severity of KOA but did not have a TKA, but this did not translate into increased physical activity or mental health.
Level of Evidence
Level III.
Keywords: osteoarthritis, pain, physical activity, physical function, total knee replacement
Abbreviations
- 20MWS
20‐m walk speed
- 5CST
five times chair‐to‐stand test
- 5STS
five times sit‐to‐stand test
- KL
Kellgren–Lawrence
- KOA
knee osteoarthritis
- KOOS
Knee Injury and Osteoarthritis Outcome Score
- MCS
Mental Component Score
- OAI
Osteoarthritis Initiative
- PASE
Physical Activity Scale for the Elderly
- PCS
Physical Component Score
- PS
propensity score
- PSM
propensity score matching
- SF‐12
12‐item Short Form Health Survey
- SMD
standardized mean difference
- TKA
total knee arthroplasty
- WOMAC
Western Ontario and McMaster Universities Osteoarthritis Index
INTRODUCTION
Knee osteoarthritis (KOA) is one of the most common joint diseases in the world and a leading cause of disability [34]. The rising prevalence of obesity and the aging population worldwide have both contributed to an increased burden of KOA [40]. Knee pain and functional disability due to KOA can significantly reduce physical function and limit physical activity levels, thus increasing the risk of other conditions such as obesity, diabetes and cardiovascular diseases [23, 26, 53]. For patients with advanced KOA whose knee symptoms do not respond to general interventions, total knee arthroplasty (TKA) is the only treatment option. While TKA effectively alleviates symptoms and enhances knee function [10, 16, 37, 50], studies have shown that the level of physical activity does not usually improve significantly after surgery [12, 33]. More importantly, studies have shown that approximately 7%–20% of patients express dissatisfaction with their post‐operative results [13, 15, 21]. There is a diversity of reasons for post‐operative dissatisfaction. Previous studies have noted that preoperative anxiety, depression and high expectations may negatively affect patients undergoing TKA, influencing their subjective function and pain perception before and 2 years after surgery [1, 27, 41]. In addition, studies have shown that persistent or recurring post‐operative pain, functional limitations and post‐operative lifestyle changes have also been identified as one of the main causes of post‐operative dissatisfaction [6, 7, 10, 22]. Previous studies mainly focused on comparing preoperative and post‐operative symptoms in patients receiving TKA [5, 31, 32]; few have evaluated whether TKA was superior to nonsurgical treatments for knee symptoms and related physical function and physical activity. In the randomized controlled trial by Skou et al. [46], patients with moderate‐to‐severe KOA were enrolled and divided into two groups for comparison. One group underwent TKA followed by a 12‐week nonsurgical treatment regimen, which included interventions such as exercise, education, dietary advice, use of insoles and pain medication. The second group received only the 12‐week nonsurgical treatment. This study demonstrated that patients in the TKA group experienced greater improvements in knee symptoms, physical function and quality of life after 12 months compared to those in the nonsurgical group. This study demonstrated that patients in the total knee replacement group experienced greater improvements in knee symptoms, physical function, and quality of life after 12 months compared to those in the nonsurgical treatment group. Subsequent analyses showed that symptomatic relief persisted in the TKA‐treated group after 24 months [47]. Moreover, it is also important to know whether the improved knee symptoms can be translated into increased physical function and activity, thereby reducing the risk of other chronic diseases [9, 19].
This study used data from a cohort of older adults with or at increased risk of KOA to compare the effect of TKA versus non‐TKA on knee symptoms, physical function and physical activity.
METHODS
Data sources
Data were from the Osteoarthritis Initiative (OAI), an open multicenter, longitudinal, prospective observational study of adults with or at increased risk of symptomatic KOA (https://nda.nih.gov/oai). Individuals aged from 45 to 79 years were recruited from four clinical sites in the United States. Major exclusion criteria included inflammatory arthritis (e.g., rheumatoid arthritis), bilateral end‐stage KOA, bilateral TKA and contraindications to 3‐T magnetic resonance imaging. All participants were scheduled for a clinical visit every 12 months since 2002, with a maximum follow‐up of 108 months. Ethical approval was obtained from the institutional review boards of all four clinical centres participating in OAI. All participants provided written informed consent.
Study design
In this study, we excluded participants who had had a TKA before cohort entry because the aim was to evaluate changes in knee symptoms and physical function and activity before and after TKA. We also excluded participants who had undergone simultaneous bilateral TKA because the development of knee symptoms may differ with unilateral TKA [11] (see flowchart in Figure 1). To ensure an adequate sample size of participants who had TKA as well as a long follow‐up period for the evaluation of the effect of TKA on the outcome measures, we prespecified the exposure period from cohort entry to the 48 months (the landmark date) [38]. Participants who had had their first TKA during the exposure period were included in the TKA group, for whom the latest annual visit before the TKA was considered baseline (Figure 2). Participants who did not have any TKA until month 48 were included in the non‐TKA group, for whom the 48‐month visit was considered baseline (Figure 2). For both the TKA and non‐TKA groups, we extracted data on participants for a maximum of 60 months.
Figure 1.
Subject selection from OAI database. n, number of participants included in the analysis; OAI, Osteoarthritis Initiative; TKA, total knee arthroplasty.
Figure 2.
Demonstrate matching of TKA and non‐TKA groups in the OAI cohort. Participants in the TKA group had their first TKA from the cohort entry (not before) to 48 months, and the baseline for this group was the clinical visit before the TKA was performed. The propensity score matching control group was selected from participants who did not conduct a TKA at or prior to the 48‐month visit. OAI, Osteoarthritis Initiative; TKA, total knee arthroplasty.
Primary outcomes
The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain and function scores were used to assess joint pain (0–20) and functional disability (0–68) in the left and right knees, with higher scores indicating more severe symptoms [8]. This was done every 12 months over 60 months.
Secondary outcomes
The Physical Activity Scale for the Elderly (PASE) was used to measure a variety of physical activities during daily life, including household chores, work, leisure activities and exercises and walking and bicycling [52]. PASE uses a scale of 0–400, with higher scores representing higher levels of physical activity [48].
Physical ability was assessed using the 20‐m walk speed (20MWS) and the five times chair‐to‐stand tests (5CST). For the 20MWS, participants were told to walk at a normal walking speed from start to finish, and the speed was measured in metres per second. For the 5CST, participants stood from a chair, keeping their arms crossed over their chest. Participants were told that the required manoeuvre was to go from a seated position with arms crossed in front of the chest and hips touching the chair to a full standing position and then sitting down again, which was recorded as a chair‐to‐stand activity [2]. These two metrics were used to assess the recovery of lower extremity strength and function of patients.
The Physical Component Score (PCS) and Mental Component Score (MCS) from the 12‐item Short Form Health Survey (SF‐12) questionnaire were used to assess the overall physical and mental health status, respectively [14]. The physical indicators in the SF‐12 scale provide a broader assessment of an individual's overall physical health than the WOMAC function indicators [8, 51].
These secondary outcomes were assessed every 24 months from the landmark date in the OAI; therefore, we evaluated change in these outcomes over 24 and 48 months.
Covariates
The selection of covariates was based on previous studies for their potential confounding effects on the symptomatic progression of KOA [28]. They were measured or recorded at baseline, including age (years) [24], sex (man/woman) [17], body mass index (BMI; kg/m2) [44], educational level, annual household income, race (White and Other) [45] and the Kellgren–Lawrence (KL) scores for both knees using x‐ray.
Data analyses
Descriptive statistics were employed to summarize the baseline characteristics of the study population. For categorical measures, the frequency and percentage (%) of knees in each group were reported. As for continuous variables, the means and standard deviations (mean ± SD) were provided. Statistical tests such as ANOVA and chi‐square were utilized to compare the groups and assess any significant differences. Using logistic regression modelling, we determined the propensity score (PS) to undergo TKA during the exposure period. Covariates entered the propensity score were sociodemographic (age, sex, BMI, education, annual household income and race), KOA severity (KL score in both knees, WOMAC pain and WOMAC function scores), and health status (SF‐12, PASE, 20MWS and 5CST) at baseline. We used callipers with a width equal to the log standard deviation of the propensity score of 0.2. The nearest‐neighbour matching method was used to match each individual in the TKA group to one individual in the non‐TKA group based on the proximity of the PS between individuals in the two groups [4].
The standardized mean difference (SMD) was measured to determine the covariate balance before propensity score matching (PSM) and after PSM with SMD > 0.1 indicating an imbalance. We developed a GEE using a Gaussian distribution and homogeneity link function to test the effect of TKA on the outcome measures, using an exchangeable correlation structure to account for within‐subject correlations. The ‘treatment’ and ‘month’ variables and their interaction were included as predictors, with adjustment for imbalanced covariates post‐PSM (i.e., SMD > 0.1). Multiple imputation by chained equations was used to address missing data caused by loss to follow‐up and nonresponses, with five imputations performed assuming missing at random.
Complete‐case analysis was performed as a sensitivity analysis. Another sensitivity analysis was conducted by excluding participants in the TKA group (and matched controls) who underwent a contralateral TKA surgery and follow‐up data on participants in the non‐TKA group from the implementation of a TKA.
All statistical analyses were performed by R software (version 4.2.1). Statistical significance was set at a p value of less than 0.05 (two‐tailed).
RESULT
Cohort characteristics
Out of the initial 4796 cohort participants, 63 had undergone TKA before cohort entry, and 17 underwent simultaneous bilateral knee replacements during the exposure period. Due to missing variables for inclusion matching at baseline, 1746 participants were excluded from the matching process. Subsequently, 86 participants experienced their first TKA during the exposure period. Following 1:1 PSM with patients who had not undergone TKA during this period, 82 pairs of participants (mean 65.24 years, 57.9% female) were selected for inclusion in our study (Figure 1).
Matching
Table 1 shows the baseline characteristics of participants in the TKA and non‐TKA groups before and after PSM. Eighty‐two participants (56% women, mean 64.8 ± 8.4 years) in the TKA group were 1:1 matched to 82 participants in the non‐TKA group. No statistically significant differences were found in the variables included for matching the TKA and non‐TKA groups after PSM.
Table 1.
Baseline characteristics of study participants before and after propensity score matching (PSM).
| Before PSM | After PSM | |||||||
|---|---|---|---|---|---|---|---|---|
| No‐TKA (N = 2868) | TKA (N = 86) | p | SMD | No‐TKA (N = 82) | TKA (N = 82) | p | SMD | |
| Age (mean (SD)) | 64.28 (8.92) | 64.64 (9.06) | 0.710 | 0.041 | 65.71 (8.39) | 64.77 (8.98) | 0.490 | 0.108 |
| Sex (%) | ||||||||
| Male | 1283 (44.49) | 36 (41.86) | 0.709 | 0.053 | 33 (40.24) | 36 (43.90) | 0.752 | 0.074 |
| Female | 1601 (55.51) | 50 (58.14) | 49 (59.76) | 46 (56.10) | ||||
| BMI (mean (SD)) | 28.471 (4.84) | 30.45 (4.69) | <0.001 | 0.415 | 30.75 (5.23) | 30.25 (4.65) | 0.520 | 0.101 |
| Education level (%) | ||||||||
| ≤High school | 365 (12.66) | 20 (23.26) | 0.008 | 0.312 | 20 (24.39) | 19 (23.17) | 0.361 | 0.224 |
| College | 1298 (45.01) | 39 (45.35) | 43 (52.44) | 36 (43.90) | ||||
| Graduate degree | 1221 (42.34) | 27 (31.40) | 19 (23.17) | 27 (32.93) | ||||
| Income (%) | ||||||||
| <25k | 306 (10.61) | 7 (8.14) | 0.907 | 0.086 | 10 (12.20) | 7 (8.54) | 0.852 | 0.139 |
| 25–50k | 710 (24.62) | 22 (25.58) | 23 (28.05) | 22 (26.83) | ||||
| 50–100k | 1093 (37.90) | 33 (38.37) | 30 (36.59) | 31 (37.80) | ||||
| >100k | 775 (26.87) | 24 (27.91) | 19 (23.17) | 22 (26.83) | ||||
| Race (%) | ||||||||
| White | 2393 (82.98) | 73 (84.88) | 0.750 | 0.052 | 68 (82.93) | 69 (84.15) | 1.000 | 0.033 |
| Other | 491 (17.02) | 13 (15.12) | 14 (17.07) | 13 (15.85) | ||||
| KL in left knee (%) | ||||||||
| ≤2 | 2365 (82.00) | 30 (34.88) | <0.001 | 1.136 | 31 (37.80) | 29 (35.37) | 0.934 | 0.058 |
| 3 | 370 (12.83) | 24 (27.91) | 23 (28.05) | 23 (28.05) | ||||
| 4 | 149 (5.17) | 32 (37.21) | 28 (34.15) | 30 (36.59) | ||||
| KL in right knee (%) | ||||||||
| ≤2 | 2342 (81.21) | 27 (31.40) | <0.001 | 1.185 | 21 (25.61) | 26 (31.71) | 0.656 | 0.144 |
| 3 | 407 (14.11) | 33 (38.37) | 32 (39.02) | 31 (37.80) | ||||
| 4 | 135 (4.68) | 26 (30.23) | 29 (35.37) | 25 (30.49) | ||||
| WOMAC Pain in left knee (mean (SD)) | 1.71 (2.85) | 5.00 (4.30) | <0.001 | 0.902 | 4.60 (4.45) | 4.76 (4.18) | 0.815 | 0.037 |
| WOMAC Pain in right knee (mean (SD)) | 1.90 (2.83) | 5.05 (3.87) | <0.001 | 0.927 | 5.65 (4.19) | 4.90 (3.79) | 0.235 | 0.186 |
| WOMAC Function in left knee (mean (SD)) | 5.66 (9.42) | 18.11 (13.38) | <0.001 | 1.076 | 16.70 (13.93) | 17.26 (12.94) | 0.791 | 0.042 |
| WOMAC Function in right knee (mean (SD)) | 5.93 (9.30) | 16.44 (13.61) | <0.001 | 0.903 | 16.67 (12.39) | 15.83 (13.34) | 0.677 | 0.065 |
| SF‐12 Mental (mean (SD)) | 54.09 (8.19) | 56.60 (6.99) | 0.005 | 0.329 | 55.39 (9.04) | 56.24 (6.95) | 0.500 | 0.106 |
| SF‐12 Physical (mean (SD)) | 48.99 (8.77) | 38.71 (9.85) | <0.001 | 1.103 | 38.84 (10.55) | 39.34 (9.48) | 0.750 | 0.050 |
| PASE (mean (SD)) | 159.56 (81.88) | 133.49 (79.93) | 0.004 | 0.322 | 139.26 (77.26) | 136.62 (80.41) | 0.831 | 0.033 |
| 20‐m speed (mean (SD)) | 1.33 (0.20) | 1.19 (0.22) | <0.001 | 0.650 | 1.18 (0.21) | 1.20 (0.22) | 0.548 | 0.094 |
| The 5‐time chair‐to‐stand tests (mean (SD)) | 10.86 (3.43) | 13.43 (4.30) | <0.001 | 0.661 | 12.40 (3.30) | 13.54 (4.28) | 0.058 | 0.299 |
Abbreviations: BMI, body mass index; KL, Kellgren–Lawrence; PASE, Physical Activity Scale for the Elderly; SD, standard deviation; SF‐12, Short Form 12‐Item Health Survey; SMD, standardized mean difference; TKA, total knee arthroplasty; WOMAC, Western Ontario and McMaster Universities arthritis index.
Primary outcome
The treatment effect of TKA on knee symptoms continued to increase until 24 months and plateaued thereafter (Pain: β = −2.09 to −3.29, p < 0.001; Function: β = −7.05 to −10.12, p < 0.001) (Figure 3). In contrast, knee symptoms in the non‐TKA group were stable throughout the study. From baseline to the 60‐month follow‐up, both knee pain and functional disability in the TKA group showed a greater improvement compared to the non‐TKA group (Table 2).
Figure 3.
Change in (a) the Western Ontario and McMaster University Index (WOMAC) pain and (b) function scores from baseline in the TKA and non‐TKA groups over 60 months. *Positive scores indicate worsening and negative improvement from baseline. Vertical bars indicate 95% CIs for the mean scores. CI, confidence interval; TKA, total knee arthroplasty.
Table 2.
Change in knee symptoms between the TKA and non‐TKA groups over 60 months.
| TKA | Non‐TKA | Between‐group difference TKA − Non‐TKA | ||||
|---|---|---|---|---|---|---|
| β (95% CI) | p | β (95% CI) | p | β (95% CI) | p | |
| WOMAC Pain (0–20) | ||||||
| 12 | −3.00 (−4.07 to −1.93) | <0.001 | −0.91 (−1.87 to 0.04) | 0.062 | −2.09 (−3.46 to −0.72) | 0.003 |
| 24 | −4.73 (−5.69 to −3.77) | <0.001 | −0.39 (−1.54 to 0.76) | 0.510 | −4.26 (−5.38 to −3.14) | <0.001 |
| 36 | −4.47 (−5.69 to −3.24) | <0.001 | −0.81 (−1.99 to 0.36) | 0.182 | −3.83 (−5.01 to −2.65) | <0.001 |
| 48 | −4.46 (−5.87 to −3.06) | <0.001 | −0.45 (−1.83 to 0.92) | 0.522 | −4.24 (−5.49 to −2.98) | <0.001 |
| 60 | −4.09 (−5.33 to −2.85) | <0.001 | −0.72 (−2.77 to 1.33) | 0.499 | −3.29 (−4.59 to −1.99) | <0.001 |
| WOMAC Function (0–68) | ||||||
| 12 | −9.84 (−13.19 to −6.49) | <0.001 | −2.05 (−5.13 to 1.03) | 0.197 | −7.05 (−11.09 to −3.01) | <0.001 |
| 24 | −14.40 (−17.51 to −11.29) | <0.001 | −0.46 (−4.41 to 3.49) | 0.819 | −12.99 (−16.70 to −9.28) | <0.001 |
| 36 | −13.42 (−17.12 to −9.72) | <0.001 | −0.59 (−5.29 to 4.12) | 0.809 | −11.85 (−15.69 to −8.01) | <0.001 |
| 48 | −13.90 (−17.98 to −9.82) | <0.001 | −0.01 (−5.31 to 5.29) | 0.998 | −12.96 (−17.19 to −8.73) | <0.001 |
| 60 | −12.81 (−17.55 to −8.07) | <0.001 | 0.06 (−8.76 to 8.88) | 0.990 | −10.12 (−14.21 to −6.03) | <0.001 |
Note: Adjust for covariates at baselines: age, BMI, education, income, KL score of the right knee, Pain scores in the right knee, mental score of SF‐12 and the five times chair‐to‐stand tests.
Abbreviations: 95% CI, 95% confidence interval; BMI, body mass index; KL, Kellgren–Lawrence; SF‐12, Short Form 12‐Item Health Survey; TKA, total knee arthroplasty; WOMAC, Western Ontario and McMaster University Osteoarthritis.
Secondary outcomes
Physical function (20MWS and 5CST) and overall physical health (SF‐12 physical health) were improved in the TKA group at 24 months and slightly regressed at 48 months (Physical: β = 4.18, 95% CI = [1.45–6.91], p = 0.003; 20MWS: β = 0.04, 95% CI = [−0.01 to 0.09], p = 0.093; 5CST: β = −1.41, 95% CI = [−2.34 to −0.48], p = 0.003) (Figure 4). Mental health score (SF‐12 mental health) and physical activity (PASE) were slightly decreased in both groups (PASE: β = 5.72, 95% CI = [−15.46 to 26.90], p = 0.597; Mental: β = 0.04, 95% CI = [−2.47 to 2.54], p = 0.976). For between‐group comparisons (Table 3), overall physical health and 5CST were significantly improved over 24 and 48 months in the TKA group, compared to the non‐TKA group, and 20MWS showed a statistically non‐significant trend to be improved in the TKA group. There were no significant between‐group differences for changes in mental health score and physical activity over 24 or 48 months (Table 3).
Figure 4.
Comparison between the TKA and non‐TKA groups on change in secondary outcome. (a) The change of SF‐12 Mental, Positive scores indicate improvement and negative indicate deterioration from baseline. (b) The change of SF‐12 Physical, Positive scores indicate improvement and negative indicate deterioration from baseline. (c) The change of PASE, Positive scores indicate improvement and negative indicate deterioration from baseline. (d) The change of 20‐m walk speed, Positive scores indicate improvement and negative indicate deterioration from baseline. (e) the change of five times chair stands, timed, Positive scores indicate worsening and negative improvement from baseline. Vertical bars indicate 95% CIs for the mean scores. 95% CI, 95% confidence interval; PASE, Physical Activity Scale for the Elderly; SF‐12, Short Form 12‐Item Health Survey; TKA, total knee arthroplasty.
Table 3.
Change in physical/mental health, physical function and physical activity between the TKA and non‐TKA groups over 48 months.
| TKA | Non‐TKA | Between‐group difference TKA − Non‐TKA | ||||
|---|---|---|---|---|---|---|
| β (95% CI) | p | β (95% CI) | p | β (95% CI) | p | |
| SF‐12 Mental (0‐100) | ||||||
| 24 | −1.48 (−4.20 to 1.25) | 0.298 | −1.93 (−5.72 to 1.86) | 0.334 | 0.04 (−2.47 to 2.54) | 0.976 |
| 48 | −2.22 (−5.66 to 1.21) | 0.223 | −2.86 (−7.08 to 1.37) | 0.205 | 0.73 (−1.92 to 3.38) | 0.589 |
| SF‐12 Physical (0‐100) | ||||||
| 24 | 5.62 (2.18–9.06) | 0.005 | −0.08 (−3.95 to 3.78) | 0.967 | 4.18 (1.45–6.91) | 0.003 |
| 48 | 4.85 (0.37–9.33) | 0.052 | −0.11 (−5.20 to 4.98) | 0.967 | 2.90 (−0.07 to 5.87) | 0.057 |
| PASE (0‐400) | ||||||
| 24 | 7.50 (−29.39 to 44.39) | 0.696 | 1.88 (−61.17 to 64.92) | 0.955 | 5.72 (−15.46 to 26.90) | 0.597 |
| 48 | −0.87 (−60.74 to 59.00) | 0.978 | −6.25 (−53.44 to 40.93) | 0.801 | −1.73 (−24.13 to 20.66) | 0.880 |
| 20‐m speed (m/s) | ||||||
| 24 | 0.06 (−0.05 to 0.17) | 0.337 | 0.01 (−0.21 to 0.23) | 0.908 | 0.04 (−0.01 to 0.09) | 0.093 |
| 48 | 0.05 (−0.16 to 0.26) | 0.672 | 0.00 (−0.16 to 0.15) | 0.971 | 0.04 (−0.01 to 0.09) | 0.114 |
| The 5‐time chair‐to‐stand tests (s) | ||||||
| 24 | −1.00 (−3.88 to 1.88) | 0.512 | −0.22 (−4.58 to 4.15) | 0.927 | −1.41 (−2.34 to −0.48) | 0.003 |
| 48 | −0.56 (−4.70 to 3.58) | 0.797 | −0.20 (−3.12 to 2.73) | 0.900 | −1.23 (−2.43 to −0.03) | 0.048 |
Note: Adjust for covariates at baselines: age, BMI, education, income, KL score of the right knee, Pain scores in the right knee, mental score of SF‐12 and the five times chair‐to‐stand tests.
Abbreviations: 95% CI, 95% confidence interval; BMI, body mass index; KL, Kellgren–Lawrence; PASE, Physical Activity Scale for the Elderly; SF‐12, 12‐item Short Form Health Survey; TKA, total knee arthroplasty.
Sensitivity analyses
Complete‐case analysis did not materially change the main findings (Table S1). During the 60‐month follow‐up, 8 participants in the TKA group underwent a contralateral TKA surgery and 17 in the non‐TKA group conducted a TKA. The exclusion of the 8 participants in the TKA group and matched controls and follow‐up data on the 17 participants in the non‐TKA group from the implementation of a TKA did not change the main findings, except that the between‐group effect of TKA on change in 20MWS over 48 months became statistically significant (Table S2).
DISCUSSION
In this study, we found that compared to the non‐TKA group, patients in the TKA group had substantially greater improvement in knee pain and knee function, physical function and overall physical health throughout the study period. However, no significant between‐group differences were found in changes in mental health and physical activity. These findings suggest a strong and long‐term effect of TKA on knee symptoms and physical function/health, but the effect has not been translated into increased physical activity, which is crucial to other chronic diseases.
In a randomized controlled trial conducted by Skou et al., TKA followed by non‐surgical treatment was more effective in improving knee pain and function over 12 months compared to non‐surgical treatment alone [46]. Post hoc analyses also indicated a significant and stable effect of TKA on knee symptoms and quality of life over 24 months [47]. Our study confirmed these findings and further indicated an effect of TKA on knee symptoms over 60 months, and the effect was not attenuated until the end of follow‐up. An explanation for these findings could be that, in the first 2 years post‐surgery, patients typically participate in various post‐operative rehabilitation protocols, including physiotherapy, rehabilitation exercises and regular follow‐up [20, 25, 36]. Such measures may alleviate post‐operative pain and facilitate joint function recovery. Thus, the gradual reduction in pain may reflect the effectiveness of these rehabilitation strategies. After 2 years, patients may reach a stable rehabilitation phase, where pain levels may reach a manageable threshold, with further improvement potentially constrained by the limits of post‐operative rehabilitation. These findings support current opinions that TKA surgery is a safe and effective treatment approach for relieving knee symptoms [43, 49].
In our study, overall physical health and 5CST were significantly improved in the TKA group at both 24 and 48 months, although a trend of diminution was observed. Moreover, the between‐group difference in 20MWS was not statistically significant in the main analyses, and this can be due to the fact that the 20MWS test only reflects a normal walking speed, which is less influenced by knee symptoms. Our findings are mostly consistent with another study evaluating the impact of TKA on gait, the five times sit‐to‐stand test (5STS), and performance‐based and self‐reported functional outcomes within one year after surgery [29]. They observed significant improvements in gait patterns, 5STS and functional scores assessed using the Knee Injury and Osteoarthritis Outcome Score (KOOS) over 1 year, although TKA‐treated patients did not reach the levels observed in the healthy control group [29]. Taken together, TKA may improve the physical function of patients with KOA, but full restoration is unlikely.
In a systematic review [3], there were only marginal changes in PASE scores at 6 months post‐TKA, with limited evidence of more substantial changes at 1 year after surgery. Our study did not demonstrate significant increases in physical activity in the TKA group at 24 and 48 months, nor were noticeable improvements in psychological well‐being observed between 24 and 48 months. Various factors may have contributed to the lack of treatment effects on mental health improvement and physical activity levels post‐operatively. Inadequate post‐operative pain management, socioeconomic disparities affecting access to rehabilitation resources, and the perceived level of social support in the post‐operative period are all potential influencers of patients' motivation to effectively engage in physical activities and manage their mental well‐being. Moreover, fear of re‐injury, the presence of comorbidities, cultural beliefs around physical activity and mental health, and the absence of individualized rehabilitation plans tailored to patients' specific needs may further contribute to the discrepancy between improved physical function and stagnant mental health scores and physical activity levels post‐operatively. It is worth noting that this inconsistency may be attributed to the rehabilitation phase commonly experienced by patients [18], during which those with greater adherence to rehabilitation plans tend to achieve more favourable physical and mental recovery outcomes [35]. Over time, however, rehabilitation progress may slow or plateau, potentially leading to decreased physical activity levels.
Additionally, patients with preexisting mental health conditions may face heightened psychological pressures post‐surgery, potentially affecting their rehabilitation process and outcomes [39]. Although improvements in physical function may occur, they do not necessarily translate to increased levels of physical activity or benefits for other chronic conditions. Concerns about compromising the stability and longevity of the implant through high‐intensity physical activities may also deter patients from engaging in such activities. Moreover, patients with high preoperative expectations may experience dissatisfaction if post‐operative outcomes fall short of their expectations [30].
Scott et al. [42] indicated an elevated incidence of clinically significant depressive symptoms pre‐ and post‐operatively among TKA patients compared to the general population. A prolonged post‐operative recovery period may heighten the psychological burden, highlighting a need for further investigation into the mental health benefits following TKA. Insufficient physical activity not only compromises physical health but also adversely impacts mental health, further reducing patients' willingness to engage in physical activity. Therefore, the provision of a comprehensive rehabilitation programme that incorporates psychological counselling and structured rehabilitation guidance is essential for helping patients overcome these barriers.
The strengths of this study include the long follow‐up period and the matched non‐TKA group of patients who had a similar severity of KOA. Our study has several limitations. First, due to the lack of specific dates of TKA procedures in the OAI cohort, participants were asked only during the annual follow‐ups if they had undergone TKA surgery within the past year. To address this limitation, we selected the 48th month as a significant time point and included patients who underwent their first TKA surgery within the 0–48 months period in the exposed group. This approach allowed us to estimate TKA exposure reasonably while maintaining methodological consistency. Second, it is noteworthy that data collection for SF‐12, PASE, 20MWS and 5CST in the OAI cohort occurred every 2 years after the landmark date. Consequently, we considered these variables as secondary outcomes, using the two‐year interval as a designated time point for subsequent assessments. This approach allowed us to further investigate the effects of TKA beyond our primary outcome measures, thereby providing additional valuable insights into patient outcomes post‐surgery. Third, the sample size of this study was relatively small, which may have limited the generalizability and representativeness of the findings. As an observational study, we have maximized the use of available data and did not conduct power analysis due to the nature of the study design. However, further larger studies are warranted to confirm our findings regarding the lack of a significant effect of TKA on mental health and physical activity over 48 months. Finally, it is important to note that PSM, while effective in reducing selection bias, cannot account for unmeasured confounding variables inherent in observational studies. Factors such as patient preferences, socioeconomic status and unrecorded comorbidities could potentially influence the outcomes and were not captured in our analysis. Despite this, the effects of TKA on WOMAC pain and function over 12–24 months in our study were comparable to those from the 1‐year RCT and its 2‐year post‐hoc analysis by Skou et al., suggesting that our findings are robust.
In conclusion, TKA substantially improved knee symptoms, physical function and overall physical health post‐operation and remained stable over 48 or 60 months, compared to those who had a similar severity of KOA but did not have a TKA, but this did not translate into increased physical activity or mental health.
AUTHOR CONTRIBUTIONS
Liru Ge: Methodology; investigation; data curation; writing—original draft; writing—review and editing. Junjie Wang: Conceptualization; investigation; formal analysis; writing—review and editing. Haonan Fang: Methodology; writing—review and editing. Yining Wang: Supervision, writing—review and editing. Ziyuan Shen: Supervision, writing—review and editing. Rui Zhu: Conceptualization; writing—review and editing. Guoqi Cai: Conceptualization; methodology; project administration; supervision; writing—review and editing.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
ETHICS STATEMENT
The ethics statement is not available.
Supporting information
Supporting information.
ACKNOWLEDGEMENTS
This work was supported by the National Natural Science Foundation of China (82103933).
Ge L, Wang J, Fang H, Wang Y, Shen Z, Cai G. Effects of total knee arthroplasty on symptoms, function and activity over 5 years in knee osteoarthritis: a propensity‐score matched study. J Exp Orthop. 2025;12:e70185. 10.1002/jeo2.70185
DATA AVAILABILITY STATEMENT
The data sets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
REFERENCES
- 1. Aalders MB, van der List JP, Keijser LCM, Benner JL. Anxiety and depression prior to total knee arthroplasty are associated with worse pain and subjective function: a prospective comparative study. Knee Surg Sports Traumatol Arthrosc. 2025;33(1):308–318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Alcazar J, Losa‐Reyna J, Rodriguez‐Lopez C, Alfaro‐Acha A, Rodriguez‐Mañas L, Ara I, et al. The sit‐to‐stand muscle power test: an easy, inexpensive and portable procedure to assess muscle power in older people. Exp Geront. 2018;112:38–43. [DOI] [PubMed] [Google Scholar]
- 3. Arnold JB, Walters JL, Ferrar KE. Does physical activity increase after total hip or knee arthroplasty for osteoarthritis? A systematic review. J Orthop Sports Phys Ther. 2016;46(6):431–442. [DOI] [PubMed] [Google Scholar]
- 4. Austin PC. Comparing paired vs non‐paired statistical methods of analyses when making inferences about absolute risk reductions in propensity‐score matched samples. Stat Med. 2011;30(11):1292–1301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Bachmeier CJM, March LM, Cross MJ, Lapsley HM, Tribe KL, Courtenay BG, et al. A comparison of outcomes in osteoarthritis patients undergoing total hip and knee replacement surgery. Osteoarthr Cartil. 2001;9(2):137–146. [DOI] [PubMed] [Google Scholar]
- 6. Baker PN, van der Meulen JH, Lewsey J, Gregg PJ. The role of pain and function in determining patient satisfaction after total knee replacement. Data from the National Joint Registry for England and Wales. J Bone Joint Surg Br. 2007;89(7):893–900. [DOI] [PubMed] [Google Scholar]
- 7. Barber‐Westin SD, Noyes FR. Aerobic physical fitness and recreational sports participation after total knee arthroplasty. Sports Health. 2016;8(6):553–560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of 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(12):1833–1840. [PubMed] [Google Scholar]
- 9. Bennell KL, Hinman RS. A review of the clinical evidence for exercise in osteoarthritis of the hip and knee. J Sci Med Sport. 2011;14(1):4–9. [DOI] [PubMed] [Google Scholar]
- 10. Beswick AD, Wylde V, Gooberman‐Hill R, Blom A, Dieppe P. What proportion of patients report long‐term pain after total hip or knee replacement for osteoarthritis? A systematic review of prospective studies in unselected patients. BMJ Open. 2012;2(1):e000435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Bohm ER, Molodianovitsh K, Dragan A, Zhu N, Webster G, Masri B, et al. Outcomes of unilateral and bilateral total knee arthroplasty in 238,373 patients. Acta Orthop. 2016;87(Suppl 1):24–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Brandes M, Ringling M, Winter C, Hillmann A, Rosenbaum D. Changes in physical activity and health‐related quality of life during the first year after total knee arthroplasty. Arthritis Care Res. 2011;63(3):328–334. [DOI] [PubMed] [Google Scholar]
- 13. Bryan S, Goldsmith LJ, Davis JC, Hejazi S, MacDonald V, McAllister P, et al. Revisiting patient satisfaction following total knee arthroplasty: a longitudinal observational study. BMC Musculoskelet Disord. 2018;19(1):423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Busija L, Ackerman IN, Haas R, Wallis J, Nolte S, Bentley S, et al. Adult measures of general health and health‐related quality of life. Arthritis Care Res. 2020;72(Suppl 10):522–564. [DOI] [PubMed] [Google Scholar]
- 15. Canovas F, Dagneaux L. Quality of life after total knee arthroplasty. Orthop Traumatol Surg Res. 2018;104(1s):41. [DOI] [PubMed] [Google Scholar]
- 16. Carr AJ, Robertsson O, Graves S, Price AJ, Arden NK, Judge A, et al. Knee replacement. Lancet. 2012;379(9823):1331–1340. [DOI] [PubMed] [Google Scholar]
- 17. Cherian JJ, O'Connor MI, Robinson K, Jauregui JJ, Adleberg J, Mont MA. A prospective, longitudinal study of outcomes following total knee arthroplasty stratified by gender. J Arthroplasty. 2015;30(8):1372–1377. [DOI] [PubMed] [Google Scholar]
- 18. Fortier LM, Rockov ZA, Chen AF, Rajaee SS. Activity recommendations after total hip and total knee arthroplasty. J Bone Joint Surg. 2021;103(5):446–455. [DOI] [PubMed] [Google Scholar]
- 19. Golightly YM, Allen KD, Caine DJ. A comprehensive review of the effectiveness of different exercise programs for patients with osteoarthritis. Phys Sportsmed. 2012;40(4):52–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Henderson KG, Wallis JA, Snowdon DA. Active physiotherapy interventions following total knee arthroplasty in the hospital and inpatient rehabilitation settings: a systematic review and meta‐analysis. Physiotherapy. 2018;104(1):25–35. [DOI] [PubMed] [Google Scholar]
- 21. Ighani Arani P, Wretenberg P, Ottosson J, W‐Dahl A. Pain, function, and satisfaction after total knee arthroplasty, with or without bariatric surgery. Obes surg. 2022;32(4):1164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kim TK, Chang CB, Kang YG, Kim SJ, Seong SC. Causes and predictors of patient's dissatisfaction after uncomplicated total knee arthroplasty. J Arthroplasty. 2009;24(2):263–271. [DOI] [PubMed] [Google Scholar]
- 23. Kim WB, Kim BR, Kim SR, Han EY, Nam KW, Lee SY, et al. Comorbidities in patients with end‐stage knee OA: prevalence and effect on physical function. Arch Phys Med Rehabil. 2019;100(11):2063–2070. [DOI] [PubMed] [Google Scholar]
- 24. Klit J, Jacobsen S, Rosenlund S, Sonne‐Holm S, Troelsen A. Total knee arthroplasty in younger patients evaluated by alternative outcome measures. J Arthroplasty. 2014;29(5):912–917. [DOI] [PubMed] [Google Scholar]
- 25. Konnyu KJ, Thoma LM, Cao W, Aaron RK, Panagiotou OA, Bhuma MR, et al. Rehabilitation for total knee arthroplasty: a systematic review. Am J Phys Med Rehabil. 2023;102(1):19–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Lee JY, Han K, McAlindon TE, Park YG, Park SH. Lower leg muscle mass relates to knee pain in patients with knee osteoarthritis. Int J Rheum Dis. 2018;21(1):126–133. [DOI] [PubMed] [Google Scholar]
- 27. Lützner C, Postler A, Beyer F, Kirschner S, Lützner J. Fulfillment of expectations influence patient satisfaction 5 years after total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2019;27(7):2061–2070. [DOI] [PubMed] [Google Scholar]
- 28. Marot V, Murgier J, Carrozzo A, Reina N, Monaco E, Chiron P, et al. Determination of normal KOOS and WOMAC values in a healthy population. Knee Surg Sports Traumatol Arthrosc. 2019;27(2):541–548. [DOI] [PubMed] [Google Scholar]
- 29. Naili JE, Iversen MD, Esbjörnsson AC, Hedström M, Schwartz MH, Häger CK, et al. Deficits in functional performance and gait one year after total knee arthroplasty despite improved self‐reported function. Knee Surg Sports Traumatol Arthrosc. 2017;25(11):3378–3386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Nam HS, Yoo HJ, Ho JPY, Kim YB, Lee YS. Preoperative education on realistic expectations improves the satisfaction of patients with central sensitization after total knee arthroplasty: a randomized‐controlled trial. Knee Surg Sports Traumatol Arthrosc. 2023;31(11):4705–4715. [DOI] [PubMed] [Google Scholar]
- 31. Neuprez A, Neuprez AH, Kaux JF, Kurth W, Daniel C, Thirion T, et al. Total joint replacement improves pain, functional quality of life, and health utilities in patients with late‐stage knee and hip osteoarthritis for up to 5 years. Clin Rheumatol. 2020;39(3):861–871. [DOI] [PubMed] [Google Scholar]
- 32. Nilsdotter AK, Toksvig‐Larsen S, Roos EM. A 5 year prospective study of patient‐relevant outcomes after total knee replacement. Osteoarthr Cartil. 2009;17(5):601–606. [DOI] [PubMed] [Google Scholar]
- 33. Nilsdotter AK, Toksvig‐Larsen S, Roos EM. Knee arthroplasty: are patients' expectations fulfilled? A prospective study of pain and function in 102 patients with 5‐year follow‐up. Acta Orthop. 2009;80(1):55–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Palazzo C, Nguyen C, Lefevre‐Colau MM, Rannou F, Poiraudeau S. Risk factors and burden of osteoarthritis. Ann Phys Rehabil Med. 2016;59(3):134–138. [DOI] [PubMed] [Google Scholar]
- 35. Papalia R, Campi S, Vorini F, Zampogna B, Vasta S, Papalia G, et al. The role of physical activity and rehabilitation following hip and knee arthroplasty in the elderly. J Clin Med. 2020;9(5):1401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Paxton RJ, Forster JE, Miller MJ, Gerron KL, Stevens‐Lapsley JE, Christiansen CL. A feasibility study for improved physical activity after total knee arthroplasty. J Aging Phys Act. 2018;26(1):7–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Price AJ, Alvand A, Troelsen A, Katz JN, Hooper G, Gray A, et al. Knee replacement. Lancet. 2018;392(10158):1672–1682. [DOI] [PubMed] [Google Scholar]
- 38. Ravi B, Croxford R, Austin PC, Lipscombe L, Bierman AS, Harvey PJ, et al. The relation between total joint arthroplasty and risk for serious cardiovascular events in patients with moderate‐severe osteoarthritis: propensity score matched landmark analysis. Br J Sports Med. 2014;48(21):1580. [DOI] [PubMed] [Google Scholar]
- 39. Ren R, Lim TY, Stern BZ, Huang HH, Poeran J, Hayden BL, et al. Relationships between preoperative mental health and improvements in patient‐reported outcomes after total hip and knee arthroplasty. J Arthroplasty. 2023;38(4):655–661.e3. [DOI] [PubMed] [Google Scholar]
- 40. Safiri S, Kolahi AA, Smith E, Hill C, Bettampadi D, Mansournia MA, et al. Global, regional and national burden of osteoarthritis 1990‐2017: a systematic analysis of the Global Burden of Disease Study 2017. Ann Rheum Dis. 2020;79(6):819–828. [DOI] [PubMed] [Google Scholar]
- 41. Scott CEH, Bugler KE, Clement ND, MacDonald D, Howie CR, Biant LC. Patient expectations of arthroplasty of the hip and knee. J Bone Joint Surg Br. 2012;94(7):974–981. [DOI] [PubMed] [Google Scholar]
- 42. Scott JE, Mathias JL, Kneebone AC. Depression and anxiety after total joint replacement among older adults: a meta‐analysis. Aging Ment Health. 2016;20(12):1243–1254. [DOI] [PubMed] [Google Scholar]
- 43. Shan L, Shan B, Suzuki A, Nouh F, Saxena A. Intermediate and long‐term quality of life after total knee replacement: a systematic review and meta‐analysis. J Bone Jt Surg. 2015;97(2):156–168. [DOI] [PubMed] [Google Scholar]
- 44. Si H, Zeng Y, Shen B, Yang J, Zhou Z, Kang P, et al. The influence of body mass index on the outcomes of primary total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2015;23(6):1824–1832. [DOI] [PubMed] [Google Scholar]
- 45. Singh JA, Lu X, Rosenthal GE, Ibrahim S, Cram P. Racial disparities in knee and hip total joint arthroplasty: an 18‐year analysis of national Medicare data. Ann Rheum Dis. 2014;73(12):2107–2115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Skou ST, Roos EM, Laursen MB, Rathleff MS, Arendt‐Nielsen L, Simonsen O, et al. A randomized, controlled trial of total knee replacement. N Engl J Med. 2015;373(17):1597–1606. [DOI] [PubMed] [Google Scholar]
- 47. Skou ST, Roos EM, Laursen MB, Rathleff MS, Arendt‐Nielsen L, Rasmussen S, et al. Total knee replacement and non‐surgical treatment of knee osteoarthritis: 2‐year outcome from two parallel randomized controlled trials. Osteoarthr Cartil. 2018;26(9):1170–1180. [DOI] [PubMed] [Google Scholar]
- 48. Smith RD, McHugh GA, Quicke JG, Dziedzic KS, Healey EL. Comparison of reliability, construct validity and responsiveness of the IPAQ‐SF and PASE in adults with osteoarthritis. Musculoskeletal Care. 2021;19(4):473–483. [DOI] [PubMed] [Google Scholar]
- 49. van Tilburg J, Rathsach Andersen M. Mid‐ to long‐term complications and outcome for morbidly obese patients after total knee arthroplasty: a systematic review and meta‐analysis. EFORT Open Rev. 2022;7(5):295–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Van Spil WE, Kubassova O, Boesen M, Bay‐Jensen AC, Mobasheri A. Osteoarthritis phenotypes and novel therapeutic targets. Biochem Pharmacol. 2019;165:41–48. [DOI] [PubMed] [Google Scholar]
- 51. Ware Jr, JE , Kosinski M, Keller SD. A 12‐item short‐form health survey: construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34(3):220–233. [DOI] [PubMed] [Google Scholar]
- 52. Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for the Elderly (PASE): development and evaluation. J Clin Epidemiol. 1993;46(2):153–162. [DOI] [PubMed] [Google Scholar]
- 53. Zheng X, Wang Y, Jin X, Huang H, Chen H, Wang Y, et al. Factors influencing depression in community‐dwelling elderly patients with osteoarthritis of the knee in China: a cross‐sectional study. BMC Geriatr. 2022;22(1):453. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting information.
Data Availability Statement
The data sets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.