Skip to main content
Download PDF

Association of physical activity trajectories over 8 years and risk of knee replacement: data from the osteoarthritis initiative

BMC Musculoskeletal Disorders volume 25, Article number: 586 (2024) Cite this article

Abstract

Background

To identify physical activity (PA) trajectories in adults with or at risk of knee osteoarthritis and to evaluate the association of PA trajectories with incident knee replacement (KR).

Methods

This study used data from the Osteoarthritis Initiative. The Physical Activity Scale for the Elderly and the KR were assessed annually from baseline to 9 years. Individuals were included if they did not undergo KR surgery at baseline and had data on PA at ≥ 1 visit before KR. Latent class growth mixture Modeling was used to identify the optimal trajectories of PA before KR. Log-binomial regression models were used to assess the association between PA trajectories and the risk of KR. Data analyses were conducted in all individuals and those with radiographic osteoarthritis (ROA) and significant knee pain (Western Ontario and McMaster Osteoarthritis Index pain score of ≥ 5 on a 0–20 scale) at baseline, respectively.

Results

Of 4731 participants (mean age 61.1 years, 58.5% female), four distinct and slightly declined PA trajectories were identified. Compared to individuals with a “Low” PA trajectory, those with “Medium-low”, “Medium-high”, or “High” PA trajectories were not significantly associated with the risk of KR (risk ratios: 0.97–1.19, all p > 0.05). Similar PA trajectories and associations with the risk of KR were observed in the subgroups of individuals with radiographic osteoarthritis and those with significant knee pain at baseline, respectively.

Conclusion

In participants with or at risk of knee osteoarthritis, PA slightly declines over time and may play no role in the risk of KR.

Peer Review reports

Introduction

The incidence and prevalence of knee osteoarthritis (KOA) have increased markedly with the rising obesity rate and aging of the population [1, 2]. In large weight-bearing joints (e.g. knee and hip), osteoarthritis (OA) is a whole joint disease involving all joint tissues (cartilage, meniscus, subchondral bone, infrapatellar fat pad, and synovial membrane) [3]. OA causes an enormous burden to the population and often leads to disability requiring surgical intervention [2] such as knee replacement (KR). Kirsten et al. found that about 7.1% of newly diagnosed OA patients would undergo KR over 2.6 years [4], and more than 95% of KR is due to knee OA [5]. With the rising prevalence of KOA, the number of KR has gradually increased [6, 7]. While KR is the only option and is highly effective for end-stage KOA, KR recipients can experience persistent pain and long-term complications [5, 8, 9].

Physical activity (PA) can reduce the risk of adverse health conditions such as falling [10], cardiovascular disease (CVD), and mortality [11, 12]. It is generally believed that PA is also beneficial for KOA [13, 14], and guidelines have consistently and strongly recommended PA and exercise for patients with KOA [15, 16]. However, a meta-analysis indicates that recreational PA is not related to the risk and progression of radiographic KOA and knee pain [17], and another study using MRI to assess the progression of KOA has even shown that ≥ 10,000 steps/day may be detrimental to knee structural changes such as bone marrow lesions (BMLs) and cartilage defects [18]. Meanwhile, previous studies have also shown the self-reported PA in the highest quartile did not affect the risk of developing OA [19] and may have a beneficial effect on knee articular cartilage [20]. In a systematic review of randomized controlled trials, high-intensity exercise showed no clinically important benefits for pain and function compared with low-intensity exercise programs [21]. Moreover, while PA generally declines with age, and KOA is disabling [22, 23], studies across different populations have consistently shown that PA levels in KOA patients are comparable to those without KOA or knee pain [24,25,26]. Previous studies have found that patients with or at risk for KOA have distinct PA trajectories, and the distinct PA patterns may have different effects on the progression of KOA [26, 27]. As an important outcome of KOA, it is unclear whether the risk of KR varies among adults with different PA trajectories. We hypothesized that high PA trajectories may be protective against the risk of KR. Therefore, the aims of this study were (1) to identify PA trajectories in adults with or at risk of KOA, and (2) to evaluate the association of PA trajectories with incident KR.

Materials and methods

Study sample

This study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [28]. The data used in this study were derived from the Osteoarthritis Initiative (OAI), a multicenter, longitudinal, prospective observational study of participants with or at an increased risk of KOA. The risk factors for inclusion were older age (> 45 years), frequent knee symptoms, regular use of medications for knee symptoms, being overweight, a history of knee injury or surgery, a family history of OA, the presence of Heberden’s nodes, and engaging in activities that involve repetitive knee bending [29]. The OAI cohort included 4,796 participants aged–45–79 years at the time of recruitment. Ethics approval was obtained from the institutional review boards of the four clinical centers (Memorial Hospital of Rhode Island, the Ohio State University, the University of Pittsburgh, and the University of Maryland/Johns Hopkins) that recruited OAI participants. All the participants provided written informed consent. A detailed description of the OAI study is available at https://nda.nih.gov/oai/about-oai.html. Data of the OAI from 0 to 96 months were used in this study, where month 0 was considered as the baseline. A total of 4731 participants who did not have knee replacement surgery before baseline and had data on PA at ≥ 1 visit before a KR were included in the study; two subgroups were identified: knee pain and radiographic OA (ROA) at baseline (Fig. 1).

Fig. 1
figure 1

Study flowchart

Full size image

Assessment of PA

PA was assessed annually from baseline to 8 years using the Physical Activity Scale for the Elderly (PASE) questionnaire [30, 31]. The PASE covers 3 domains of activity: leisure activities, household activities, and occupational activities. The period covered by PASE is the past 7 days. The frequency, duration, and intensity of activity over the past seven days were recorded, and the total PASE score was calculated, which ranged from 0 to 400 or more (the maximum score was 556 in this population, with higher scores indicating greater PA. For each participant, data on PASE used for the evaluation of PA trajectories were restricted to visits before any KR was conducted.

Assessment of incident KR

Facts and dates of KR were self-reported and adjudicated from radiographs or medical records at the baseline and each follow-up visit. An outcome event was considered when KR surgery was performed during follow-up.

Assessment of knee pain and radiographic OA

Knee pain was assessed using the Western Ontario and McMaster Osteoarthritis Index (WOMAC) pain subscale over the past seven days [32]. Each item of the WOMAC pain subscale was evaluated using a five-point Likert scale (0 none, 1 mild, 2 moderate, 3 severe, 4 extreme), and the knee pain score was the sum of the individual item scores (0–20), with a higher number representing worse symptoms. Participants were considered to have significant knee pain at baseline if the WOMAC pain score was ≥ 5 [33, 34].

Knee radiographs were obtained using a fixed flexion knee radiograph, and the Kellgren-Lawrence (KL) grade was evaluated. Two projects in the OAI (Projects 15 and 37/42) were conducted to measure the KL grade, with two OAI experts blinded to each other’s reading and all other data. We used the ‘worst’ KL grade from the two projects to merge the duplicated data, as recommended in the OAI handbook. Participants were considered to have ROA at baseline if the KL grade of any knee was ≥ 2 [35].

Covariates

Covariates were selected based on previous studies [24, 25, 36]. We used baseline data on age, sex (male, female), body mass index (BMI, kg/m2), race (white, black, other), marital status (married or similar, widowed, divorced, separated, never married), level of education (< high school, high school graduate, some college, college graduate, some graduate school, and graduate degree) and income, history of knee injury (yes/no) and knee surgery (yes/no), use of medications (non-steroidal anti-inflammatory drugs [NSAIDs], steroids, painkillers and acetaminophen), physical function (SF-12 physical score), KL grade (the worse KL grade in both knees ), WOMAC pain score (the worse score in both knees). To account for the potential impact of seasonality on PA, we further adjusted for the month of the year that evaluated PA at baseline.

Statistical analysis

First, we estimated Latent Class Growth Mixture Modeling (LCGMM) using the lcmm package, a mixed model for multivariate longitudinal outcomes using a maximum likelihood estimation method, to reveal the latent class and predicted trajectories of PA [37]. They can be used to identify latent subgroups, classes, or clusters of individuals based on their common growth trajectories over time. The GRoLTS-checklist was used to ensure that these analyses are reproducible [38]. Individuals who completed at least one PASE measurement were included in the analysis. To determine the optimal number of trajectories, we tested models with different numbers and forms (linear, quadratic, and cubic) of the potential trajectories [26, 39]. The final number of trajectories was established according to the following criteria: (a) Akaike information criterion (AIC) [40], Bayesian information criterion (BIC) [41], and sample size adjusted Bayesian information criterion (SABIC) [42], with smaller values indicating a better model fit; (b) entropy (range from 0 to 1), with higher values indicating a better class separation; (c) each growth trajectory was assigned at least 5% of the total population; (d) the mean posterior probability of each trajectory class was greater than 70% [43]; (e) the interpretability and research significance of the identified trajectories. We established PA trajectories for all participants, participants with knee pain at baseline, and participants with ROA at baseline.

Second, we evaluated the associations of PA trajectories with the risk of KR using log-binomial regression models, and the results are shown as risk ratios (RR) with 95% confidence intervals (CI). Adjustment was made for age, sex, BMI, race, level of education and income, knee injury, knee surgery, use of medications, physical function, KL grade, and WOMAC pain scores. If the log-binomial regression models failed to converge, the RR was estimated using Poisson distribution and robust standard errors (SEs) [44]. The interactions between PA trajectories and BMI or sex at baseline were assessed. The results are reported as RRs and 95% CI. These analyses were also conducted in participants with significant knee pain (WOMAC pain score ≥ 5) and ROA at baseline as sensitivity analyses. We conducted another sensitivity analysis restricting participants with data on PA at 2 or more visits.

Missing data on covariates were addressed using multiple imputations with chained equations (MICE) [45]. Ten imputations were conducted using complete covariates and non-missing values of the predictor and outcome measures at baseline, assuming random missing values.

All analyses were performed using R, version 4.2.3 [46]. A two-sided p-value less than 0.05 was considered statistically significant.

Results

Participants

Of the 4731 participants included in this study, 2766 (58%) were female, the mean (standard deviation) age was 61.05 (9.17) years, 1468 (31%) had significant knee pain (i.e. WOMAC pain score ≥ 5), and 2512 (53.1%) had ROA at baseline. Table 1 shows the baseline characteristics of the participants. The time point and the number of KR can be seen in Supplementary Table 5.

Table 1 Baseline characteristics of study participants with different trajectories of physical activity
Full size table

PA trajectories

Participants were separated into four different latent classes, estimated and observed mean trajectories of PA Fig. 2A, and the predicted mean growth curves of the four distinct trajectories are shown in Fig. 2B. These were identified and labeled as “Low level” (n = 1635, 34.56%), “Medium-low level” (n = 1863, 39.38%), “Medium-high level” (n = 971, 20.52%), “High level” (n = 262, 5.54%). PA trajectories in participants with significant knee pain or ROA were similar (Supplementary Fig. 1). Compared to participants with Low levels of PA, those who had Medium-low, Medium-high, or High-level PA were more likely to be younger and have higher income and education levels (Table 1). The four trajectories had a high mean posterior probability, all higher than 0.8, indicating a good model fit (Supplementary Table 1). Similarly, trajectory analyses also showed a good model fit in participants with significant knee pain and those with ROA at baseline (Supplementary Tables 2 and Supplementary Table 3). The characteristics of PASE scores over time are reported in Supplementary Table 4.

Fig. 2
figure 2

Estimated and observed mean trajectory in total participants (Fig. 2A) and the predicted mean trajectories in total participants (Fig. 2B). Solid lines show class-specific mean predicted levels as a function of 8 years estimated from the best fitting growth mixture model (4-class linear latent class growth mixture modeling), shaded areas indicate estimated 95% confidence intervals

Full size image

Association of PA trajectories with risk of KR

Among the 4731 participants, 417 underwent KR during the 9-year follow-up period. Compared to the “Low level” PA trajectory, individuals with a trajectory of “Medium-low level” PA (adjusted RR, 1.15, 95% CI, 0.91–1.57), “Medium-high level” PA (adjusted RR, 1.07, 95% CI, 0.73–1.59), or “High level” PA (adjusted RR, 0.97, 95% CI, 0.51–1.84) were not associated with the risk of KR (Table 2).

Table 2 Association between physical activity trajectories and knee replacement
Full size table

Sensitivity analysis

Missing data on covariates ranged from 0.02–6.99% (Supplementary Table 6). Complete case analyses did not materially alter the main findings (Supplementary Table 7). PA trajectories in participants with significant knee pain or ROA were similar, and there were 3 distinct PA trajectories for both subgroups: “Low-level”, “Medium level”, and High level” (Supplementary Fig. 1). No significant associations were found between PA trajectories and KR risk in participants with significant knee pain or ROA (Table 2). In participants with 2 or more PA assessments, PA trajectories and their association with the risk of KR were similar to the main findings (Supplementary Fig. 2 and Supplementary Table 8).

Discussion

In middle-aged and older adults with or at an increased risk of KOA, four distinct slightly declined PA trajectories (High, Medium-high, Medium-low, and Low) were identified over the 8-year follow-up period, but PA trajectories may play no role in the risk of KR. Similar PA trajectories and association with the risk of KR were also observed in subgroups of participants with significant knee pain and those with ROA at baseline. These results suggest that PA trajectories are similar in adults with or at risk for KOA and that High-level PA does not increase the risk of progression to end-stage KOA.

PA levels within the four trajectories only slightly declined during the eight years, indicating that substantial changes in the average level of PA were rare over a long period in participants with or at an increased risk of KOA. We found that most (> 70%) participants showed a Low- or Medium-low level PA trajectory and that more (80%) participants with ROA and significant knee pain at baseline showed a Low- or Moderate-level PA trajectory. This is consistent with the finding that, among the general population, approximately 80% of adults and adolescents in the US are insufficiently active [47]. This may be due to many reasons, such as pain, catastrophizing, and injury [48]. A previous study showed that when patients catastrophize pain in the morning, they would engage in less PA and more sedentary behavior throughout the day [48].

Several studies have investigated the association between PA and the risk of KR, with inconsistent findings. In a meta-analysis that combined three case-control studies, the odds of KR in runners were lower than those in the controls [49]. In contrast, Wang et al. used data from a large population-based cohort study showing that PA, especially vigorous activity, increased the risk of KR [50]. Munugoda et al. also found that every 1000 steps/day increase in ambulatory activity, as measured by pedometers, was associated with a 9% greater risk of KR [51]. Nonetheless, Skou et al. indicated that PA, assessed by the PASE score, was not significantly associated with the risk of KR with knee pain at both short-term (2–2.5 years) and long-term (7 years) follow-ups [52]. However, these studies focused on the association between PA at one time point and the risk of KR, which failed to capture the dynamic characteristics of PA. The present study identified the PA trajectories of participants and explored their association with the risk of KR. We found that a Higher-level PA trajectory was not associated with a statistically significant incident KR, irrespective of whether the participants had ROA or significant knee pain at baseline. This study evaluated PA using PASE, which cannot well distinguish between activities that involve heavy, repeated load of the joints and those that do not, thus we were not able to evaluate the type of activity on the risk of TKR in this study. However, a previous study has summarized relevant evidence showing that the risk of KR is increased in those who conduct repetitive high-impact sports like soccer, team handball, and ice hockey, but not normal exercises like jogging, gymnastics, and swimming.

PA is essential for the management of KOA by controlling weight and strengthening muscles, thereby reducing the burden on the joints and the risk of knee injury or pain. Most patients consider KR surgery when they have worse radiographic KOA, serious chronic knee pain, stiffness, and functional impairment that significantly impair their quality of life. Strong evidence indicates that increased levels of PA can decrease knee pain in adults with OA [53]. Doré et al. in an MRI study showed a detrimental effect of high-intensity PA on cartilage loss [18], but this effect may be transient and cannot overweigh the beneficial effect of PA [54]. In the present study, individuals with higher PA levels did not have significantly increased odds of incident KR, even after adjusting for the ability to perform PA. The Physical Activity Guidelines for Americans indicate that both aerobic and muscle-strengthening PA are beneficial, and moving more and sitting less will benefit nearly everyone [47]. PA counseling should act as part of the standard care for individuals at high risk of KOA, especially at an early stage when PA engagement is more attainable.

The strengths of this study include the large sample size with long-term follow-up and the use of at least one PASE visit to avoid uncertain estimates in latent classes. Our study had several limitations. First, PA was assessed using a self-reported PASE questionnaire, which could be subject to reporting bias. Accelerometers can provide a more precise estimation of PA assessment but would be costly, time-consuming, and burdensome to participants. Second, assessments of PA were implemented annually, which may have missed some variations in these subjective measures. Third, as this was an observational study, residual confounding cannot be excluded. However, we included multiple confounders, especially those that may influence the implementation of KR, based on the previous literature. Fourth, PASE cannot well distinguish between activities that involve heavy, repeated load of the joints and those that do not, thus we were not able to evaluate the type of activity on the risk of TKR in this study.

In conclusion, among 4731 participants with or at high risk of KOA, four distinct and slightly declined PA trajectories were identified, with more than 80% of participants having low- or medium-level PA over 8 years. Participants with different PA levels had a similar risk of KR.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Abbreviations

AIC:

Akaike information criterion

BIC:

Bayesian information criterion

BMI:

Body mass index

BMLs:

Bone marrow lesions

CI:

Confidence intervals

CVD:

Cardiovascular disease

KL:

Kellgren-Lawrence

KR:

Knee replacement

LCGMM:

Latent class growth mixture Modeling

NSAIDs:

Non-steroidal anti-inflammatory drugs

OAI:

Osteoarthritis Initiative

PA:

Physical activity

PASE:

Physical Activity Scale for the Elderly

PP:

Posterior probability

ROA:

Radiographic osteoarthritis

RR:

Risk ratios

SABIC:

Size-adjusted Bayesian information criterion

SEs:

Standard errors

STROBE:

Strengthening the Reporting of Observational Studies in Epidemiology

WOMAC:

Western Ontario and McMaster Osteoarthritis Index

References

  1. Safiri S, Kolahi AA, Smith E, Hill C, Bettampadi D, Mansournia MA, Hoy D, Ashrafi-Asgarabad A, Sepidarkish M, Almasi-Hashiani A, 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):819–28.

    Article  PubMed  Google Scholar 

  2. Litwic A, Edwards MH, Dennison EM, Cooper C. Epidemiology and burden of osteoarthritis. Br Med Bull. 2013;105:185–99.

    Article  PubMed  Google Scholar 

  3. Katz JN, Arant KR, Loeser RFJJ. Diagnosis and treatment of hip and knee osteoarthritis: a review. 2021, 325(6):568–78.

  4. Leyland KM, Judge A, Javaid MK, Diez-Perez A, Carr A, Cooper C, Arden NK, Prieto-Alhambra D. Obesity and the relative risk of knee replacement surgery in patients with knee osteoarthritis: a prospective cohort study. Arthritis Rheumatol (Hoboken NJ). 2016;68(4):817–25.

    Article  Google Scholar 

  5. Price AJ, Alvand A, Troelsen A, Katz JN, Hooper G, Gray A, Carr A, Beard D. Knee replacement. Lancet (London England). 2018;392(10158):1672–82.

    Article  PubMed  Google Scholar 

  6. Culliford DJ, Maskell J, Beard DJ, Murray DW, Price AJ, Arden NK. Temporal trends in hip and knee replacement in the United Kingdom: 1991 to 2006. J bone Joint Surg Br Volume. 2010;92(1):130–5.

    Article  CAS  Google Scholar 

  7. Patel A, Pavlou G, Mújica-Mota RE, Toms AD. The epidemiology of revision total knee and hip arthroplasty in England and Wales: a comparative analysis with projections for the United States. A study using the National Joint Registry dataset. bone Joint J 2015, 97–b(8):1076–1081.

  8. Weinstein AM, Rome BN, Reichmann WM, Collins JE, Burbine SA, Thornhill TS, Wright J, Katz JN, Losina E. Estimating the burden of total knee replacement in the United States. J Bone Joint Surg Am. 2013;95(5):385–92.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Singh JA, Jensen MR, Harmsen WS, Gabriel SE, Lewallen DG. Cardiac and thromboembolic complications and mortality in patients undergoing total hip and total knee arthroplasty. Ann Rheum Dis. 2011;70(12):2082–8.

    Article  PubMed  Google Scholar 

  10. Sherrington C, Fairhall N, Kwok W, Wallbank G, Tiedemann A, Michaleff ZA, Ng C, Bauman A. Evidence on physical activity and falls prevention for people aged 65 + years: systematic review to inform the WHO guidelines on physical activity and sedentary behaviour. Int J Behav Nutr Phys Act. 2020;17(1):144.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Jeong SW, Kim SH, Kang SH, Kim HJ, Yoon CH, Youn TJ, Chae IH. Mortality reduction with physical activity in patients with and without cardiovascular disease. Eur Heart J. 2019;40(43):3547–55.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Dempsey PC, Rowlands AV, Strain T, Zaccardi F, Dawkins N, Razieh C, Davies MJ, Khunti KK, Edwardson CL, Wijndaele K, et al. Physical activity volume, intensity, and incident cardiovascular disease. Eur Heart J. 2022;43(46):4789–800.

    Article  PubMed  Google Scholar 

  13. Wang C, Schmid CH, Iversen MD, Harvey WF, Fielding RA, Driban JB, Price LL, Wong JB, Reid KF, Rones R, et al. Comparative effectiveness of Tai Chi Versus physical therapy for knee osteoarthritis: a Randomized Trial. Ann Intern Med. 2016;165(2):77–86.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Lo GH, Vinod S, Richard MJ, Harkey MS, McAlindon TE, Kriska AM, Rockette-Wagner B, Eaton CB, Hochberg MC, Jackson RD, et al. Association between walking for Exercise and symptomatic and structural progression in individuals with knee osteoarthritis: data from the Osteoarthritis Initiative Cohort. Arthritis Rheumatol (Hoboken NJ). 2022;74(10):1660–7.

    Article  Google Scholar 

  15. Kolasinski SL, Neogi T, Hochberg MC, Oatis C, Guyatt G, Block J, Callahan L, Copenhaver C, Dodge C, Felson D, et al. 2019 American College of Rheumatology/Arthritis Foundation Guideline for the Management of Osteoarthritis of the Hand, hip, and Knee. Arthritis Care Res. 2020;72(2):149–62.

    Article  Google Scholar 

  16. Bannuru RR, Osani MC, Vaysbrot EE, Arden NK, Bennell K, Bierma-Zeinstra SMA, Kraus VB, Lohmander LS, Abbott JH, Bhandari M, et al. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthritis Cartilage. 2019;27(11):1578–89.

    Article  CAS  PubMed  Google Scholar 

  17. Gates LS, Perry TA, Golightly YM, Nelson AE, Callahan LF, Felson D, Nevitt M, Jones G, Cooper C, Batt ME, et al. Recreational physical activity and risk of incident knee osteoarthritis: an International Meta-Analysis of Individual participant-Level Data. Arthritis Rheumatol (Hoboken NJ). 2022;74(4):612–22.

    Article  CAS  Google Scholar 

  18. Doré DA, Winzenberg TM, Ding C, Otahal P, Pelletier JP, Martel-Pelletier J, Cicuttini FM, Jones G. The association between objectively measured physical activity and knee structural change using MRI. Ann Rheum Dis. 2013;72(7):1170–5.

    Article  PubMed  Google Scholar 

  19. Felson DT, Niu J, Yang T, Torner J, Lewis CE, Aliabadi P, Sack B, Sharma L, Guermazi A, Goggins J, et al. Physical activity, alignment and knee osteoarthritis: data from MOST and the OAI. Osteoarthritis Cartilage. 2013;21(6):789–95.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Racunica TL, Teichtahl AJ, Wang Y, Wluka AE, English DR, Giles GG, O’Sullivan R, Cicuttini FMJAC. Effect of physical activity on articular knee joint structures in community-based adults. Research. 2007;57(7):1261–8.

  21. Regnaux JP, Lefevre-Colau MM, Trinquart L, Nguyen C, Boutron I, Brosseau L, Ravaud P. High-intensity versus low-intensity physical activity or exercise in people with hip or knee osteoarthritis. Cochrane Database Syst Rev. 2015;2015(10):Cd010203.

  22. Troiano RP, Berrigan D, Dodd KW, Mâsse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc, 40(1):181–8.

  23. American College of Sports, Chodzko-Zajko M, Proctor WJ, Fiatarone Singh DN, Minson MA, Nigg CT, Salem CR, Skinner GJ. American College of Sports Medicine position stand. Exercise and physical activity for older adults. Med Sci Sports Exerc. 2009;41(7):1510–30.

    Article  Google Scholar 

  24. White DK, Tudor-Locke C, Zhang Y, Niu J, Felson DT, Gross KD, Nevitt MC, Lewis CE, Torner J, Neogi T. Prospective change in daily walking over 2 years in older adults with or at risk of knee osteoarthritis: the MOST study. Osteoarthritis Cartilage. 2016;24(2):246–53.

    Article  CAS  PubMed  Google Scholar 

  25. Thoma LM, Dunlop D, Song J, Lee J, Tudor-Locke C, Aguiar EJ, Master H, Christiansen MB, White DK. Are older adults with symptomatic knee osteoarthritis less active than the General Population? Analysis from the Osteoarthritis Initiative and the National Health and Nutrition Examination Survey. Arthritis Care Res. 2018;70(10):1448–54.

    Article  Google Scholar 

  26. Chang AH, Lee JJ, Chmiel JS, Almagor O, Song J, Sharma L. Association of Long-term strenuous physical activity and extensive sitting with Incident Radiographic knee osteoarthritis. JAMA Netw Open. 2020;3(5):e204049.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Christensen JC, Jakiela JT, White DK. Factors Associated with trajectories of Physical Activity over 8 years in knee osteoarthritis. Eur J Rheumatol. 2023;10(1):1–7.

    Article  PubMed  PubMed Central  Google Scholar 

  28. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of Observational studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453–7.

    Article  Google Scholar 

  29. Nevitt M, Felson D. Lester GJPftcs: The osteoarthritis initiative. 2006, 1:2.

  30. 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):153–62.

    Article  CAS  PubMed  Google Scholar 

  31. Washburn RA, McAuley E, Katula J, Mihalko SL, Boileau RA. The physical activity scale for the elderly (PASE): evidence for validity. J Clin Epidemiol. 1999;52(7):643–51.

    Article  CAS  PubMed  Google Scholar 

  32. 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 Rhuematol. 1988;15(12):1833–40.

    CAS  Google Scholar 

  33. Bellamy N, Bell MJ, Pericak D, Goldsmith CH, Torrance GW, Raynauld JP, Walker V, Tugwell P, Polisson R. BLISS index for analyzing knee osteoarthritis trials data. J Clin Epidemiol. 2007;60(2):124–32.

    Article  PubMed  Google Scholar 

  34. Wang Y, Teichtahl AJ, Abram F, Hussain SM, Pelletier JP, Cicuttini FM, Martel-Pelletier J. Knee pain as a predictor of structural progression over 4 years: data from the Osteoarthritis Initiative, a prospective cohort study. Arthritis Res Therapy. 2018;20(1):250.

    Article  Google Scholar 

  35. Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957;16(4):494–502.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Cai G, Li X, Zhang Y, Wang Y, Ma Y, Xu S, Shuai Z, Peng X, Pan F. Knee symptom but not radiographic knee osteoarthritis increases the risk of falls and fractures: results from the Osteoarthritis Initiative. Osteoarthritis Cartilage. 2022;30(3):436–42.

    Article  CAS  PubMed  Google Scholar 

  37. Proust-Lima C, Philipps V, Liquet B. Estimation of extended mixed models using latent classes and latent processes: the R Package lcmm. J Stat Softw 2017, 78(2).

  38. Van De Schoot R, Sijbrandij M, Winter SD, Depaoli S, Vermunt JKJSEMAMJ. The GRoLTS-checklist: guidelines for reporting on latent trajectory studies. Struct Equation Modeling: Multidisciplinary J. 2017;24(3):451–67.

    Article  Google Scholar 

  39. Pahor M, Guralnik JM, Ambrosius WT, Blair S, Bonds DE, Church TS, Espeland MA, Fielding RA, Gill TM, Groessl EJ, et al. Effect of structured physical activity on prevention of major mobility disability in older adults: the LIFE study randomized clinical trial. JAMA. 2014;311(23):2387–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Akaike HJItoac. A new look at the statistical model identification. 1974, 19(6):716–23.

  41. Schwarz GJTaos: Estimating the dimension of a model. 1978:461–464.

  42. Sen S, Cohen AS, Kim S-HJA. Model selection for multilevel mixture Rasch models. 2019, 43(4):272–89.

  43. Strauss VY, Jones PW, Kadam UT, Jordan KPJJ. Distinct trajectories of multimorbidity in primary care were identified using latent class growth analysis. 2014, 67(10):1163–71.

  44. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–6.

    Article  PubMed  Google Scholar 

  45. Zhang Z. Multiple imputation with multivariate imputation by chained equation (MICE) package. Annals Translational Med. 2016;4(2):30.

    Google Scholar 

  46. R. A Language and Environment for Statistical Computing [https://www.R-project.org/].

  47. Piercy KL, Troiano RP, Ballard RM, Carlson SA, Fulton JE, Galuska DA, George SM, Olson RD. The physical activity guidelines for americans. JAMA. 2018;320(19):2020–8.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Zhaoyang R, Martire LM, Darnall BD. Daily pain catastrophizing predicts less physical activity and more sedentary behavior in older adults with osteoarthritis. Pain. 2020;161(11):2603–10.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Timmins KA, Leech RD, Batt ME, Edwards KL. Running and knee osteoarthritis: a systematic review and Meta-analysis. Am J Sports Med. 2017;45(6):1447–57.

    Article  PubMed  Google Scholar 

  50. Wang Y, Simpson JA, Wluka AE, Teichtahl AJ, English DR, Giles GG, Graves S, Cicuttini FM. Is physical activity a risk factor for primary knee or hip replacement due to osteoarthritis? A prospective cohort study. J Rhuematol. 2011;38(2):350–7.

    Article  CAS  Google Scholar 

  51. Munugoda IP, Wills K, Cicuttini F, Graves SE, Lorimer M, Jones G, Callisaya ML, Aitken D. The association between ambulatory activity, body composition and hip or knee joint replacement due to osteoarthritis: a prospective cohort study. Osteoarthritis Cartilage. 2018;26(5):671–9.

    Article  CAS  PubMed  Google Scholar 

  52. Skou ST, Wise BL, Lewis CE, Felson D, Nevitt M, Segal NA, Multicenter Osteoarthritis Study G. Muscle strength, physical performance and physical activity as predictors of future knee replacement: a prospective cohort study. Osteoarthr Cartil. 2016;24(8):1350–6.

    Article  CAS  Google Scholar 

  53. Kraus VB, Sprow K, Powell KE, Buchner D, Bloodgood B, Piercy K, George SM, Kraus WE. Effects of physical activity in knee and hip osteoarthritis: a systematic Umbrella Review. Med Sci Sports Exerc. 2019;51(6):1324–39.

    Article  PubMed  PubMed Central  Google Scholar 

  54. Coburn SL, Crossley KM, Kemp JL, Warden SJ, West TJ, Bruder AM, Mentiplay BF, Culvenor AG. Is running good or bad for your knees? A systematic review and meta-analysis of cartilage morphology and composition changes in the tibiofemoral and patellofemoral joints. Osteoarthritis Cartilage. 2023;31(2):144–57.

    Article  CAS  PubMed  Google Scholar 

Download references

Funding

This work was supported by the National Natural Science Foundation of China (82103933). This article was prepared using an Osteoarthritis Initiative (OAI) public-use data set, and its contents do not necessarily reflect the opinions or views of the OAI Study Investigators, the NIH, or the private funding partners of the OAI. The OAI is a public–private partnership between the NIH (contracts N01-AR-2-2258, N01-AR-2-2259, N01-AR-2-2260, N01-AR-2-2261, and N01-AR-2-2262) and private funding partners (Merck Research Laboratories, Novartis Pharmaceuticals, GlaxoSmithKline, and Pfizer, Inc.) and is conducted by the OAI Study Investigators. Private sector funding for the OAI is managed by the Foundation for the NIH. The authors of this article are not part of the OAI investigative team.

Author information

Authors and Affiliations

  1. Department of Epidemiology and Biostatistics, School of Public Health, Anhui Medical University, Hefei, Anhui, 230032, China

    Yining Wang, Ziyuan Shen, Xing Xing, Liru Ge, Faming Pan & Guoqi Cai

  2. The Inflammation and Immune-Mediated Diseases Laboratory of Anhui Province, Anhui Medical University, 81 Meishan Road, Hefei, Anhui, 230032, China

    Faming Pan

  3. Menzies Institute for Medical Research, University of Tasmania, Hobart, TAS, 7000, Australia

    Guoqi Cai

Authors
  1. Yining Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ziyuan Shen
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xing Xing
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Liru Ge
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Faming Pan
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Guoqi Cai
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

GC had full access to all the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. Conception and design: YW and GC. Data clean: YW and ZS. Data analysis and interpretation: YW, ZS and XX. Drafting of the paper: YW, ZS, XX, LG. Revising it critically for intellectual content: GC, FP. All authors approval of the version to be published and agree to be accountable for all aspects of the work.

Corresponding author

Correspondence to Guoqi Cai.

Ethics declarations

Ethics approval and consent to participate

Ethics approvals were obtained from the institutional review board at each of the four clinical centers (Memorial Hospital of Rhode Island, the Ohio State University, the University of Pittsburgh, and the University of Maryland/Johns Hopkins) that recruited OAI participants. All participants provided written informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wang, Y., Shen, Z., Xing, X. et al. Association of physical activity trajectories over 8 years and risk of knee replacement: data from the osteoarthritis initiative. BMC Musculoskelet Disord 25, 586 (2024). https://doi.org/10.1186/s12891-024-07710-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12891-024-07710-9

Keywords

BMC Musculoskeletal Disorders

ISSN: 1471-2474

Contact us