The University of Texas Health Science Center at San Antonio, Department of Radiology, 7703 Floyd Curl Drive, MC 7800, San Antonio, TX 78229-3900. E-mail: firstname.lastname@example.org
To evaluate the association of baseline frequent knee bending activities with the prevalence and progression of cartilage and meniscal abnormalities over 3 years and to assess the effect of frequent knee bending on the different knee compartments with 3T magnetic resonance imaging (MRI).
We studied 115 subjects without radiographic knee osteoarthritis (OA) but with risk factors for OA from the Osteoarthritis Initiative database. The inclusion criteria at baseline were age 45–55 years, body mass index of 19–27 kg/m2, Western Ontario and McMaster Universities Osteoarthritis Index pain score of 0, and Kellgren/Lawrence grade <2. Knee bending activities (kneeling, squatting, stair climbing, and weight lifting) were assessed by questionnaire at the baseline clinic visit. Cartilage and meniscal abnormalities were graded using the Whole-Organ MRI Score. Logistic regression was used to determine the association of frequent knee bending with cartilage and meniscal abnormalities.
Frequent knee bending activities were associated with an increased risk of prevalent cartilage lesions (odds ratio [OR] 3.63, 95% confidence interval [95% CI] 1.39–9.52), in particular in the patellofemoral compartment (OR 3.09, 95% CI 1.22–7.79). The increase in risk was higher in subjects involved in ≥2 knee bending activities. At 3-year followup, individuals reporting frequent knee bending were more likely to show progression of cartilage damage (OR 4.12, 95% CI 1.27–13.36) and meniscal abnormalities (OR 4.34, 95% CI 1.16–16.32).
Frequent knee bending activities were associated with a higher prevalence of knee cartilage lesions (particularly in the patellofemoral compartment) and with an increased risk of progression of cartilage and meniscal lesions in asymptomatic middle-aged subjects.
Osteoarthritis (OA) of the knee is one of the major causes of functional limitation and physical disability worldwide. The overall lifetime risk of symptomatic knee OA has been found to be as high as 40–50% (). The etiology of knee OA is multifactorial. Excessive loading across the knee joint is considered an important risk factor in the pathogenesis of knee OA ([2, 3]). Activities during which the knee is flexed beyond 90° are known to place a greater load across the entire knee joint and potentially cause more damage to the articular cartilage ([4-6]). Several studies have reported that the risk of knee OA is increased by work that involves prolonged bending of the knees ([7-9]). These studies, however, have relied on job title as an index of presumed work place activity, rather than directly monitoring the particular types of repetitive knee movements ([5, 10]). Furthermore, it is increasingly recognized that for the given activity, the tibiofemoral and patellofemoral joints are exposed to different loads and contact stresses ([4-6]). Previous epidemiologic studies have mostly evaluated knee OA based on radiographs, which are insensitive to early changes in cartilage morphology, and are not well suited for identification of the location where cartilage is most affected ([7, 8, 11-19]). Magnetic resonance imaging (MRI)–based studies can directly visualize structural morphology in all compartments of the knee; therefore, it plays a key role in assessing the severity of OA and in monitoring its progression. Understanding how frequent knee bending activities may affect cartilage and menisci by observing morphologic change on MRI may provide further insights into the role of biomechanical loading in the pathogenesis of OA in different compartments of the knee joint.
Although many previous studies have evaluated subjects with symptomatic and radiographic OA ([10, 11, 14, 20]), the present study focused on subjects who are at risk for developing OA, but in knees that do not have radiographic evidence or pain. Since early morphologic degeneration in the joint may not be detected using radiography ([21, 22]), this study used MRI to assess cartilage and meniscal abnormalities. Moreover, we classified subjects with frequent knee bending using questionnaires that assessed certain types of physical activity directly. Therefore, the goals of our study were to analyze the association between repetitive knee bending activity and focal knee abnormalities both cross-sectionally and longitudinally over a period of 3 years, and also to examine separately the impact of knee bending activities on the different compartments of the knee.
Box 1. Significance & Innovations
Frequent knee bending is associated with a higher prevalence and an increased risk of progression of cartilage and meniscal lesions in asymptomatic middle-aged subjects.
Behavior modification may play an important role in delaying onset and progression of osteoarthritis.
MATERIALS AND METHODS
Data used in this study were obtained from the publically accessible Osteoarthritis Initiative (OAI) database (http://www.oai.ucsf.edu/). The OAI is a multicenter, longitudinal study aimed at assessing biomarkers of OA, including those derived from MRI, and their impact on the natural evolution of joint degeneration. The OAI provides a cross-sectional and longitudinal data set that includes both MRI and radiographic images of 4,796 subjects scanned annually over 4 years. We used clinical and imaging data sets at baseline (0.2.2 and 0.E.1) and at 36 months of followup (5.2.1 and 5.E.1). This work has received the approval of the OAI Publications Committee based on a review of its scientific content and data interpretation.
A total of 115 subjects were selected from the incidence subcohort of the OAI, aiming to exclude subjects who have knee symptoms and/or radiographic knee OA. The OAI defined a participant as having knee OA if radiographs of one or both knees showed a Kellgren/Lawrence (K/L) grade ≥2 combined with pain or stiffness in that knee on most days of the month for the last year. Participants in the incidence subcohort of the OAI did not have symptomatic, radiographic knee OA in either knee, but had risk factors for developing knee OA (e.g., obesity, history of knee injury or knee surgery, a family history of total knee replacement, or Heberden's nodes). The flow diagram provided in Figure 1 details the process of subject selection from the OAI. We excluded participants with K/L grade ≥2 and also participants with knee pain. Further exclusion criteria for the OAI included rheumatoid arthritis, bilateral severe knee joint space narrowing, and a positive pregnancy test. The specific inclusion criteria for this study were age 45–55 years, body mass index (BMI) of 19–27 kg/m2, Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain score of 0 in both knees (), and K/L score ≤1 () in the right knee at baseline. These criteria were chosen in order to examine a middle-aged, nonobese, asymptomatic population without radiographic evidence of OA.
WOMAC and Physical Activity Scale for the Elderly (PASE) questionnaires
The WOMAC is a well-established clinical tool used to assess clinical symptoms of OA, including pain, stiffness, and physical and emotional function (). In this study, a WOMAC pain score of 0 was used to exclude patients with knee pain in the 7 days prior to the baseline MRI. The PASE is a well-established questionnaire to quantify physical activity in older individuals and was used to quantify physical activity levels in all subjects ().
Repetitive knee bending exposure
Using a questionnaire on repetitive knee bending activity, subjects were asked at their initial eligibility interview, “As part of your daily activities, either at work or outside of work, do you currently do any of the following physical activities on most days? These activities do not have to be done all at once. They can be done over the course of the day.” Among the physical activities queried were 1) climbing up a total of 10 or more flights of stairs, 2) kneeling for 30 minutes or more, 3) squatting or deep knee bending for 30 minutes or more, and 4) moving objects weighing 25 pounds or more by hand. We categorized participants into the frequent knee bending group if they reported exposure to at least one of these activities.
Radiographic knee assessment
Baseline weight-bearing posteroanterior bilateral knee radiographs with fixed flexion technique (20–30° of flexion and 10° of internal rotation) were obtained (). Two readers (WV and LN) quantified the OA findings of the right knee in consensus using the K/L grading scale ().
MRIs were obtained using 4 identical 3.0T scanners (Siemens Magnetom Trio) at 4 clinical sites taking part in this study (Ohio State University, Columbus, Ohio; University of Maryland, Baltimore, Maryland; University of Pittsburgh, Pittsburgh, Pennsylvania; and Brown University, Pawtucket, Rhode Island). The OAI MRI protocol has been described in depth previously (). For the analyses performed in this study, we used specific, standard MRI sequences (); briefly, the following sequences were used: a sagittal 2-dimensional intermediate-weighted fast spin-echo sequence with fat suppression (repetition time [TR] 3,200 msec, echo time [TE] 30 msec, spatial resolution 0.357 mm − 0.511 mm, slice thickness 3.0 mm), a coronal 2-dimensional intermediate-weighted fast spin-echo sequence (TR 3,700 msec, TE 29 msec, spatial resolution 0.365 mm − 0.456 mm, slice thickness 3.0 mm), a sagittal 3-dimensional dual-echo steady state sequence (TR 16.3 msec, TE 4.7 msec, spatial resolution 0.365 mm − 0.456 mm, slice thickness 0.7 mm), and a 3-dimensional fast low-angle shot sequence with selective water excitation (TR 20 msec, TE 7.57 msec, spatial resolution 0.313 mm − 0.313 mm, slice thickness 1.5 mm).
MRIs of the right knee were reviewed on picture archiving and communication system workstations (Agfa). A board-certified radiologist with 5 years of experience with musculoskeletal imaging (WV) and a fourth-year radiology resident (LN) reviewed all of the images independently and graded cartilage and meniscal abnormalities for each image in total at the same time using a modified semiquantitative Whole-Organ Magnetic Resonance Imaging Score (WORMS) ([29, 30]), as outlined below. The readers were blinded to knee bending status and clinical and radiographic data, and read the paired images with knowledge of the time sequence.
The weighted kappa statistics of the WORMS grade for intrarater reliabilities for reader 1 (WV) and reader 2 (LN) were 0.98 and 0.95 for cartilage score grading and 0.99 and 0.94 for meniscal score grading, respectively. The interrater reliabilities for cartilage and meniscal score grading were 0.90 and 0.91, respectively. In case of disagreement, a consensus reading was performed with a musculoskeletal radiologist with more than 20 years of experience (TML). All readers were blinded to the frequent knee bending status of the subjects.
Cartilage morphology grading
Since only relatively mild cartilage abnormalities were expected, we analyzed only 6 rather than the 15 compartments of the original WORMS. This included the patella, trochlea, medial and lateral femur, and medial and lateral tibia. Decreasing the number of compartments could have potentially affected the number of grade 4 and grade 6 lesions, which, however, are expectedly rare in this patient population. Cartilage signal and morphology were scored using an 8-point scale, where 0 = normal thickness and signal; 1 = normal thickness but increased signal on fluid-sensitive (intermediate-weighted) sequences; 2 = partial-thickness focal defect <1 cm in greatest width; 2.5 = full-thickness focal defect <1 cm in greatest width; 3 = multiple areas of partial-thickness (grade 2) defects intermixed with areas of normal thickness, or a partial-thickness defect wider than 1 cm but <75% of the region; 4 = diffuse (≥75% of the region) partial-thickness loss; 5 = multiple areas of full-thickness loss (grade 2.5) or a full-thickness defect wider than 1 cm but <75% of the region; and 6 = diffuse (≥75% of the region) full-thickness loss.
Meniscal morphology scoring
Meniscal morphology was assessed in 6 regions, i.e., the medial and lateral sides of the anterior, body, and posterior region; an additional grade was added to the meniscal WORMS classification (intrasubstance degeneration) to better assess early degenerative disease. The grading scale ranged from 0–4, where 0 = normal, 1 = intrasubstance abnormalities, 2 = nondisplaced tear, 3 = displaced or complex tear, and 4 = complete destruction and maceration of the meniscus.
Statistical analysis was performed using Stata software, version 11 (StataCorp). Descriptive statistics for baseline demographic data (i.e., age, sex, BMI), PASE (), and K/L score were calculated for each group. Continuous variables are shown as the mean ± SD and compared using the independent-sample t-test. Categorical variables are shown as the frequency (percentage) and compared using the chi-square test.
Since WORMS is not a true ordinal or continuous scale, we analyzed the prevalence and progression outcomes as binary; 3 separate outcomes of cartilage abnormalities were analyzed to examine the association with frequent knee bending activity: 1) the prevalence of cartilage lesions, defined as WORMS ≥1; 2) the prevalence of cartilage defects, defined as WORMS ≥2; and 3) the progression of cartilage morphologic abnormalities, defined as an increase in any subregion of a compartment of the WORMS from baseline to the 36-month followup. The outcomes for the menisci were analyzed in the same way as those for the cartilage: 1) the prevalence of meniscal lesions, defined as WORMS ≥1; 2) the prevalence of meniscal tear, defined as WORMS ≥2; and 3) the progression of meniscal morphologic abnormalities, defined as an increase in the WORMS. Cartilage inhomogeneity and intrasubstance abnormality of the menisci (WORMS ≥1) were used as an outcome in order to detect the earliest sign of morphologic changes; however, these findings were more likely subjective, and therefore we used morphologic cartilage defect and meniscal tear (WORMS ≥2) as another outcome for evaluation of cartilage and menisci.
Differences in outcomes between the groups were assessed using bivariate and multivariate logistic regression and are reported as the odds ratio (OR) and 95% confidence interval (95% CI). The multivariate logistic models were adjusted for the predefined potential confounders, including age, sex, BMI, history of knee injury, and history of knee surgery. All P values were derived from the likelihood ratio test. For the analysis of progression in each compartment, because of the low number of progressing lesions after 3 years, exact logistic regression was used and only sex was entered into the model. A P value of less than 0.05 was considered to be statistically significant.
Among the 115 subjects in this study, 55 were men and 60 were women. The mean ± SD age of the participants was 50.8 ± 2.9 years, while their mean ± SD BMI was 24.1 ± 1.9 kg/m2. The number of subjects reporting frequent knee bending activities at baseline, including stair climbing, kneeling, squatting, heavy lifting, or none of these activities, was 62 (53.9%), 18 (15.7%), 20 (17.4%), 41 (35.7%), and 31 (27.0%), respectively. Forty subjects reported exposure to only one kind of these activities and 42 subjects reported exposure to ≥2 kinds of these activities. Characteristics of the subjects categorized by their reported exposure to frequent knee bending activity are shown in Table 1. No significant differences in terms of sex, age, BMI, PASE, K/L score, and other OA risk factors were found between the 2 groups.
Table 1. Baseline characteristics of the 115 subjects from the Osteoarthritis Initiative incidence cohort by frequent knee bending activities*
BMI = body mass index; PASE = Physical Activity Scale for the Elderly; OA = osteoarthritis; K/L = Kellgren/Lawrence.
aPearson's chi-square test for categorical variables and an independent-sample t-test for continuous variables were used.
Male sex, no. (%)
Age, mean ± SD years
50.7 ± 2.9
51.3 ± 2.9
BMI, mean ± SD kg/m2
24.2 ± 1.8
23.7 ± 1.9
Baseline PASE, mean ± SD
207.8 ± 84.5
185.6 ± 76.6
Other OA risk factors, no. (%)
Previous knee injury
Baseline K/L score 1, no. (%)
Prevalence of cartilage and meniscal lesions at baseline
Frequent knee bending activities were associated with an increased risk of prevalent cartilage lesions (WORMS ≥1) in the whole knee, as shown in Table 2. We also found a trend of increased risk for cartilage defects (WORMS ≥2); however, this did not reach statistical significance. In additional exploratory analyses by compartment of the knee, we observed a significant association between frequent knee bending and cartilage lesions in the patellofemoral compartment (Table 2). For prevalent meniscal abnormality, there was no significant association between frequent knee bending activity and increased risk of meniscal abnormality (Tables 2 and 3).
Table 2. Risk for cartilage and meniscal abnormalities at baseline according to frequent knee bending activities (n = 115) for any abnormality and morphologic abnormalities*
Table 3. Risk for prevalent cartilage and meniscal abnormalities at baseline by number of frequent knee bending activities involved*
No. of frequent knee bending activities involved, OR (95% CI)a
P for trend
Two groups were separated with exposure to 1 knee bending activity (n = 40) and ≥2 knee bending activities (n = 42), such as frequent kneeling and squatting. OR = odds ratio; 95% CI = 95% confidence interval; WORMS = Whole-Organ Magnetic Resonance Imaging Score.
aAdjusted for sex, age, body mass index, history of knee injury, and knee surgery.
bStatistically significant difference between subjects performing 1 frequent knee bending activity and those performing no frequent knee bending activities.
cStatistically significant difference between subjects performing ≥2 frequent knee bending activities and those performing no frequent knee bending activities.
Participation in more than one knee bending activity was associated with higher ORs of having cartilage abnormalities, as shown in Table 3. The OR of having cartilage lesions in the subjects who performed only one type of frequent knee bending was 3.08 (95% CI 1.01–9.35) and for subjects involved in ≥2 types of frequent knee bending activities was 4.28 (95% CI 1.37–13.35). The OR of having cartilage defects in the subjects who performed only one type of frequent knee bending was 1.33 (95% CI 0.50–3.55) and for subjects involved in ≥2 types of frequent knee bending was 3.35 (95% CI 1.23–9.06).
We explored whether there are differences by types of knee bending activity and found that each activity was again associated with increased risk for cartilage abnormalities, but the ORs were no longer statistically significant.
Progression of cartilage and meniscal scores at 3-year followup
Thirty-three subjects (28.7%) showed progression of the cartilage score in at least one compartment (Figure 2), whereas 29 subjects (25.2%) had progression of the medial or lateral meniscal score. The progression of cartilage and meniscal scores after 3-year followup by frequent knee bending is shown in Table 4. Risk for progression of cartilage morphology scores in any compartment of the knee joint was greater in the group with frequent knee bending (OR 4.12, 95% CI 1.27–13.36). Analyzing the regions separately, we were not able to identify a significant progression of cartilage in any regions with a limited number of outcomes. For menisci, we found significant worsening of the morphology scores overall (OR 4.34, 95% CI 1.16–16.32), as well as for the medial meniscus (OR 7.38, 95% CI 1.06–321.62).
Table 4. Comparison of the progression of focal knee abnormalities at 3-year followup of 115 subjects with and without repetitive knee bending activity*
Values are the number (percentage) unless otherwise indicated. OR = odds ratio; 95% CI = 95% confidence interval; PF = patellofemoral; TF = tibiofemoral.
aMultivariable logistic regression adjusted for sex, age, body mass index, history of knee injury, and knee surgery was used for overall progression, and exact logistic regression adjusted for sex only was used for each region.
bStatistically significant difference between the frequent knee bending group and the no frequent knee bending group.
Using data from the OAI study, we found that frequent knee bending activity in asymptomatic middle-aged subjects without radiographic OA in the studied knee but who had risk factors for OA is associated with an increased risk of knee cartilage abnormality, especially at the patellofemoral compartment. The results of this study suggest that increased biomechanical loading at the knee during bending activities may have a detrimental effect on cartilage. Deep knee bending is known to dramatically increase the stresses and the loads in the knee, with knee flexion to 90° resulting in tibiofemoral joint stress of up to 26.6 MPa, which exceeds the threshold at which cartilage damage occurs (). Knee bending activities such as squatting have also been shown to increase sheer stresses on the patellofemoral joint ().
Cooper et al proposed 2 mechanisms whereby repetitive knee use might increase the risk of OA (): occupations characterized by prolonged periods of kneeling and squatting may increase the risk of meniscal or ligamentous damage to the knee, and such lesions are known risk factors for knee OA, and repetitive loading might directly induce cartilage loss. Our observation that the risk of cartilage and meniscal changes from knee bending is independent of that from self-reported knee injury and knee surgery suggests a direct effect. Most other studies that have investigated the role of occupational activities at the knee have used radiographs to examine individuals with knee OA ([8, 13, 14]). However, radiographic studies cannot discriminate early abnormalities in cartilage and have limitations in evaluating the changes in the patellofemoral compartment.
Although conventional radiography did not demonstrate joint space narrowing or significant osteophytes in the knees we studied, MRI detected cartilage and meniscal pathology and disease progression over a relatively short observation period of 3 years. The patella cartilage had the highest prevalence of abnormalities compared to the other compartments in our study (Figure 3). Previous studies have also demonstrated that MRI is superior to radiography in demonstrating OA-related morphologic lesions, irrespective of symptoms (). Therefore, soft tissue degeneration in the knee may not closely correspond with joint space narrowing, and radiography may not be optimal for assessing early-stage joint degeneration.
Two recent cross-sectional studies used MRI to evaluate cartilage change in relation to certain types of occupational activities, including knee bending, and examined the changes in different compartments ([20, 33]). One study in a male population with knee OA found that occupational exposure to frequent squatting/kneeling and heavy lifting was associated with deleterious patellofemoral cartilage changes (). The other study enrolled asymptomatic adult women and found an increased risk for prevalent patella cartilage defects in subjects whose vocational tasks necessitated frequent knee flexion (). Our results, which showed an increased risk of prevalent patellofemoral cartilage lesions with frequent knee bending, are consistent with these previous studies.
The natural history of OA varies greatly. Generally, OA develops progressively over several years. Although we found an association of progression of the overall cartilage score at the 3-year followup period in subjects involved in frequent knee bending, we were not able to show significant results for the individual compartments separately. We also found an association of frequent knee bending and progression of the overall meniscal score and medial meniscal abnormality. The MRI-based study by Rytter et al showed that the prevalence of degenerative tears was higher in the medial meniscus among floor layers, who engaged in frequent kneeling work positions, compared to that found in graphic designers (). They proposed that high medial contact forces and the relative immobility of the posterior part of the medial meniscus combined with its larger size could explain the result of an unbalanced distribution of abnormality between the medial and lateral menisci.
The strength of our study was that we evaluated focal knee abnormalities using MRI, which allows detecting early changes of the cartilage and evaluation of changes in each compartment of the knee separately. In addition to examining the prevalence of morphologic change in a cross-sectional study, we evaluated the changes in a longitudinal 3-year followup period. In our study, we evaluated asymptomatic subjects, which helped to eliminate recall bias that might occur in subjects with knee OA who are more likely to report their history of frequent knee bending.
Our study has some potential limitations. Although we used the activity questionnaire to identify exposure to squatting, kneeling, climbing stairs, and heavy lifting, we did not have information on the duration of these activities of interest to assess the potential influence of time of activity on the results. Also, we adjusted for history of knee injury or surgery based on information obtained from self-reported questionnaires. Since the number of morphologically progressive lesions over a 3-year period in our study was not large, we were not able to show any significant results in the compartment-specific analysis. These findings will need to be confirmed in larger longitudinal studies and/or with longer followup. Another limitation of our study was that we did not correlate the development of morphologic abnormalities with the development of symptoms. The discordance between clinical and radiographic knee OA is well known; therefore, there is a need to further investigate the correlation of structural abnormalities within the knee with symptoms and functional abnormalities.
The statistical analysis used for this study performed multiple comparisons (such as across different compartments of the knee, medial and lateral menisci, etc.). The higher number of comparisons performed can lead to an increased chance of false-positive results. This is an additional limitation of this study.
In conclusion, frequent knee bending activities are associated with an increased risk of prevalence and progression of cartilage abnormalities, and with an increased risk of worsening meniscal damage, in asymptomatic middle-aged subjects at risk for OA, particularly at the patellofemoral joint. Given the concern that these early lesions may progress to an advanced stage of OA, our findings emphasize the importance of determining which particular knee bending activities, and at what frequency, increase the risk of cartilage and meniscus damage in order to develop behavioral modification and intervention for the preventive management of OA.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Alizai had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Virayavanich, Alizai, Baum, Nardo, Nevitt, Lynch, McCulloch, Link.
Acquisition of data. Virayavanich, Alizai, Nardo, Nevitt, Lynch, McCulloch, Link.
Analysis and interpretation of data. Virayavanich, Alizai, Baum, Nardo, Nevitt, Lynch, McCulloch, Link.
ROLE OF THE STUDY SPONSOR
Merck Research Laboratories, Novartis Pharmaceuticals Corporation, GlaxoSmithKline, and Pfizer, Inc. had no role in the study design, data collection, data analysis, or writing of the manuscript, as well as approval of the content of the submitted manuscript. Publication of this article was not contingent on the approval of the sponsors.