Longitudinal effect of vigorous physical activity on patella cartilage morphology in people without clinical knee disease
There are few data concerning possible long-term effects of physical activity on cartilage change in the patellofemoral compartment. We examined the effect of participation in vigorous physical activity on changes to patella cartilage over 2 years.
A total of 297 healthy adults ages 50–79 years with no history of knee injury or symptoms were recruited from an existing study. Physical activity data were obtained by questionnaire at baseline (2003–2004). Patella cartilage volume and defects were determined by magnetic resonance imaging at baseline (2003–2004) and followup (2006–2007).
Participation in vigorous physical activity at baseline was associated with a reduced rate of patella cartilage volume loss (−21.2 mm3 per annum [95% confidence interval (95% CI) −41.5, −1.0; P = 0.04]) and a trend toward less risk of worsening patella cartilage defects (odds ratio 0.4; 95% CI 0.2, 1.1 [P = 0.07]) over the subsequent 2 years. In the subgroup with no significant patella cartilage defects at baseline (n = 192), participation in vigorous physical activity was associated with a reduced annual rate of patella cartilage volume loss (95% CI −53.8, −7.8; P = 0.03) and a trend for fewer new patella cartilage defects (95% CI 0.1, 1.1; P = 0.08). No significant relationships were found between vigorous physical activity and cartilage volume change or defect progression in the subgroup with prevalent patella cartilage defects at baseline.
These observations suggest that vigorous physical activity is beneficial to patellofemoral joints for people without preexisting cartilage damage. Weight-bearing vigorous physical activity might, therefore, be useful in the prevention of patellofemoral osteoarthritis.
The knee is a tricompartmental joint, comprising the medial and lateral tibiofemoral and patellofemoral compartments. Although most epidemiologic studies examining the health of the knee joint have focused on the tibiofemoral compartments, the patellofemoral compartment represents a common source of pain and disability (1, 2). Radiographic evidence of patellofemoral osteoarthritis (OA) is present in approximately one-third of people age >60 years (1).
Determining modifiable factors that expedite or delay deleterious changes to joint structure prior to the development of clinical OA may help target preventive strategies for patellofemoral OA. One such modifiable factor may be physical activity. Studies have confirmed the importance of physical activity to the development of cartilage in children (3), and have demonstrated the rapid cartilage loss that occurs following forced immobility (e.g., after spinal cord injury) in adults (4, 5). Although these findings suggest that physical activity is required for cartilage development and maintenance at the knee, it is unclear whether longer-term exposure to physical activity is beneficial or detrimental to adult articular cartilage, particularly at the patellofemoral joint.
The paucity of data concerning the possible associations of physical activity with the patellofemoral joint may be due in part to limitations inherent in using radiographs to indirectly examine cartilage using the joint space width as a surrogate measure (6, 7). Using magnetic resonance imaging (MRI), it is now possible to directly examine articular cartilage for both those with OA and healthy subjects prior to clinical disease (8–12). Both cartilage volume and surface lesions known as cartilage defects can be validly, reliably, and sensitively measured using MRI (13). Although cartilage volume represents a geometric measurement of the entire cartilage plate and is known to be lost at an accelerated rate during the development of knee OA (14), cartilage defects are an independent and early structural feature of degenerative change that increases the risk of subsequent cartilage loss and eventual joint replacement surgery (15–17).
The aim of this cohort study was to examine the associations between baseline vigorous physical activity and any subsequent change to patella cartilage morphology over 2 years in middle-aged people without clinical knee OA.
SUBJECTS AND METHODS
Study subjects were recruited from an existing cohort: the Melbourne Collaborative Cohort Study (MCCS), which is a prospective cohort study of 41,528 people ages 40–69 years at recruitment (1990–1994) with the aim of examining the role of lifestyle and genetic factors in the risk of cancer and chronic diseases from middle age and older (18). Because our intent was to investigate subjects with no significant current or past knee disease, individuals were excluded if, in the last 5 years, they had knee pain lasting >24 hours, a previous knee injury requiring non–weight-bearing treatment for >24 hours or surgery (including arthroscopy), or a history of any arthritis diagnosed by a medical practitioner. A further exclusion criterion was a contraindication to MRI. We invited subjects who fulfilled our inclusion criteria and attended the first year of round 3 of followup of the MCCS, which began in 2003. We used quota sampling, whereby recruitment ceased when our target sample of ∼300 subjects was achieved. The study was approved by The Cancer Council Victoria Human Research Ethics Committee and the Standing Committee on Ethics in Research Involving Human Subjects of Monash University. All participants gave written informed consent.
Assessment of physical activity.
At the time subjects were recruited into the present study (2003–2004), physical activity was assessed on the basis of activity in the previous 7 days. Vigorous activity was assessed by asking, “Over the past 7 days, how often did you engage in strenuous sport and recreational activities such as jogging, swimming, cycling, singles tennis, aerobic dance, skiing or other similar activities?”, “What were these activities?”, and “How often per week?”, where a subject could chose from never, seldom (1 to 2 days), sometimes (3 to 4 days), and often (5–7 days). From these data, it could be determined whether the form of vigorous physical activity was weight bearing (e.g., jogging, tennis, dance, etc.) or non–weight bearing (e.g., swimming, hydrotherapy).
Weight was measured with bulky clothing removed using electronic scales. Height was measured using a stadiometer with shoes removed. Body mass index (BMI) was calculated (weight [kg]/height [m2]). The 1990–1994 data were termed baseline.
MRI and the measurement of cartilage changes.
Between 2003 and 2004, and again between 2006 and 2007, each subject underwent MRI of their dominant knee, defined by the extremity the subject first stepped off from when initiating gait. Imaging was performed in the morning to negate any potential influence that diurnal variation may have on articular cartilage. Immediately prior to MRI, each subject performed similar physical activity, with normal walking preceding the scan, followed by rested non–weight bearing. Knees were imaged in the sagittal plane on a 1.5T whole-body magnetic resonance unit (Philips 1.5T Intera; Philips Medical Systems, Eindhoven, The Netherlands) using a commercial transmit–receive extremity coil, with a T1-weighted fat-suppressed 3-dimensional gradient-recall acquisition as previously described (19, 20).
Patella cartilage volume was determined by image processing on an independent workstation using the Osiris software (University Hospital of Geneva, Geneva, Switzerland). Contours were drawn around the patella in images 1.5 mm apart on sagittal views. The assessor was blinded to both the pairing and segmentation of the patellofemoral cartilage volumes over the course of the study. The intraobserver coefficient of variation was 2.1% for patella cartilage volume (21). The reproducibility of patella cartilage volume measurements when the same subject was imaged a second time, ∼1 week later, yielded a coefficient of variation of 2.6%.
The annual change of patella cartilage volume was determined from the equation:
Cartilage defects were graded on the MRIs with a classification system that has been previously described for each compartment of the knee joint (20, 22), where grade 0 = normal cartilage; grade 1 = focal blistering and an intracartilaginous low signal intensity area with an intact surface and bottom; grade 2 = irregularities on the surface or bottom and a loss of thickness of <50%; grade 3 = deep ulceration with a loss of thickness of >50%; and grade 4 = full-thickness cartilage wear with exposure of the subchondral bone. A defect grade of >1 indicated a prevalent defect and was used to create subgroups of people with and without patella cartilage defects at baseline. Intraobserver and interobserver reliability assessed in 50 MRIs (expressed as the intraclass correlation coefficient) were 0.94 for patella cartilage volume. Change in defect grade was determined by subtracting the 2003–2004 defect grade from the 2006–2007 defect grade. A change in the defect grade of ≥1 represented a worsening of the severity of cartilage defects. If no significant patella cartilage defects were present at baseline (grade ≤1), a worsening followup grade represented an incident and clinically significant defect. If significant cartilage defects were present at baseline (i.e., a baseline grade >1), an increase in defect grade represented cartilage defect progression.
Change in patella cartilage volume was initially assessed for normality before being regressed against the physical activity variables. Logistic regression was used to examine a worsening in patella cartilage defect grade, since this was classified as a dichotomous outcome. Known confounders were adjusted for in the regression models, including baseline age, sex, BMI, and the respective 2003–2004 measure of joint structure. P values less than 0.05 were considered to be statistically significant. All analyses were performed using the SPSS statistical package, standard version 15.0 (SPSS, Chicago, IL).
The characteristics of the subjects at baseline are shown in Table 1. A total of 271 (91.2%) of the original 297 subjects completed the longitudinal MRI component. Twenty-six subjects, consisting of 9 men and 17 women, were lost to followup because of death (n = 3), poor health (n = 4), withdrawal of consent (n = 10), development of a contraindication to MRI (pacemaker; n = 4), and inability to make contact (n = 5). The subjects lost to followup tended to have a higher BMI (P = 0.06) than those who completed the study, but were not significantly different in any other variables (Table 1).
Table 1. Subject characteristics
|Age at baseline, mean ± SD years||57.8 ± 5.2||60.0 ± 7.5||0.16†|
|Women, no. (%)||169 (62)||17 (65)||0.76‡|
|Body mass index, mean ± SD kg/m2||25.7 ± 4.1||27.9 ± 5.4||0.06†|
|Participation in vigorous physical activity, no. (%)||89 (33)||8 (31)||0.83‡|
|Weekly frequency of vigorous physical activity in the previous 7 days|| || || |
| Never||182||18|| |
| Seldom (1–2 days)||49||5|| |
| Sometimes (3–4 days)||23||1|| |
| Often (5–7 days)||17||2|| |
|Participation in vigorous weight-bearing physical activity only, no. (%)||72 (27)||4 (15)||0.14‡|
|Participation in vigorous non–weight-bearing physical activity only, no. (%)||17 (6)||4 (15)||0.23‡|
|Baseline patella cartilage volume, mean ± SD mm3||2,666 ± 893||2,555 ± 811||0.52†|
There was a significant reduction in patella cartilage volume observed between baseline and followup (P < 0.0001), with a mean ± SD annual loss of patella cartilage volume of 49.6 ± 78.8 mm3, or 1.8% per annum (95% confidence interval [95% CI] 1.3, 2.9%). Thirty-seven subjects (12.4%) demonstrated worsening patella cartilage defects. The mean ± SD time between baseline and followup analyses was 2.3 ± 0.4 years.
A total of 182 subjects had not participated in vigorous physical activity in the prior 7 days, whereas 89 had. Subjects that had participated in vigorous physical activity tended to be younger and were more likely to be male, have a lower BMI, and a greater baseline patella cartilage volume, as well as a smaller annual change in patella cartilage volume, than those who had not participated in vigorous physical activity (Table 2). The mean ± SEM patella cartilage volume loss for those participating in physical activity (32.5 ± 8.4 mm3) was significantly less than the amount lost by those who did not participate in vigorous physical activity (56.5 ± 5.9 mm3) after adjustment for age, sex, and BMI (P = 0.03 for the difference). These relationships persisted when a subgroup excluding people who had participated in non–weight-bearing vigorous physical activity (n = 17) was excluded (data not shown).
Table 2. Subject characteristics based on participation in vigorous physical activity*
|Age at baseline, years||57.0 ± 5.4||58.2 ± 5.1||0.06‡|
|Women, no. (%)||47 (53)||122 (67)||0.02§|
|Body mass index, kg/m2||24.9 ± 3.5||26.1 ± 4.3||0.02‡|
|Baseline patella cartilage volume, mm3||3,090 ± 999||2,834 ± 902||0.04‡|
|Annual change in patella cartilage volume, mm3||32.5 ± 78.8||58.0 ± 77.6||0.01‡|
Participation in vigorous physical activity at baseline was associated with a significant reduction in loss of annual patella cartilage volume before (P = 0.001) and after adjustment for potential cofounders (β = −23.8; 95% CI −44.1, −3.5 [P = 0.02]; whereby a negative regression coefficient represents a reduction in the annual loss of patella cartilage volume) (Table 3). There was a trend for a dose response in benefit, with increasing frequency (in days per week) of vigorous physical activity being associated with a reduction in annual loss of patella cartilage volume before (P = 0.02) and a similar tendency after adjustment for potential confounders (β = −9.8; 95% CI −20.7, −1.1 [P = 0.08]) (Table 3).
Table 3. Relationship between participation in and weekly frequency of vigorous physical activity (2003–2004) and the change in knee cartilage morphology between 2003 and 2004 and between 2006 and 2007 (n = 271)
|Annual patella cartilage volume change, mm3†|| || || || |
| Participation in vigorous physical activity (yes/no)||−27.4 (−47.4, −7.5)||< 0.01||−23.8 (−44.1, −3.5)||0.02|
| Weekly frequency of vigorous physical activity‡||−12.4 (−22.9, −1.9)||0.02||−9.8 (−20.7, −1.1)||0.08|
|Worsening of patella cartilage defects, yes/no§|| || || || |
| Participation in vigorous physical activity (yes/no)||0.5 (0.2, 1.1)||0.07||0.4 (0.2, 1.1)||0.07|
| Weekly frequency of vigorous physical activity‡||0.6 (0.4, 1.1)||0.10||0.6 (0.4, 1.1)||0.08|
After adjustment for potential cofounders, there was a trend for participation in vigorous physical activity at baseline to be associated with a reduced risk for worsening of patella cartilage defects (odds ratio [OR] 0.4; 95% CI 0.2, 1.1 [P = 0.07]). This relationship also tended to be associated with weekly frequency (in days per week) of vigorous physical activity (OR 0.6; 95% CI 0.4, 1.1 [P = 0.08]) (Table 3). These relationships tended to persist when a subgroup excluding people who had participated in non–weight-bearing vigorous physical activity (n = 17) was excluded (data not shown).
Subgroup analyses were performed to determine whether baseline participation in vigorous physical activity was associated with the progression of preexisting patella cartilage defects or the development of incident patella cartilage defects. The characteristics of the subgroups are shown in Table 4. The severity of cartilage defects in the subgroup that had patella cartilage defects present at baseline (n = 79) were as follows: grade 2, 54% (n = 43), grade 3, 24% (n = 19), and grade 4, 22% (n = 17). Seventeen (22%) of 79 people with incident patella cartilage defects at baseline demonstrated progression of patella cartilage defects at followup. Twenty (10%) of 192 people without patella cartilage defects at baseline developed incident patella cartilage defects at followup.
Table 4. Subject characteristics based on subgroups*
|Age, years||57.7 ± 5.0||58.1 ± 4.7||0.70‡||58.1 ± 5.4||56.0 ± 4.1||0.04‡|
|Women, no. (%)||45 (73)||14 (82)||0.41§||95 (55)||16 (80)||0.10§|
|Body mass index, kg/m2||27.4 ± 4.3||25.9 ± 4.4||0.25‡||25.3 ± 3.8||24.1 ± 4.6||0.27‡|
|Participation in vigorous physical activity, no. (%)||17 (27)||4 (24)||0.75§||63 (37)||4 (20)||0.14§|
|Patella cartilage volume at baseline, mm3||2,562 ± 912||2,255 ± 845||0.20‡||3,140 ± 911||2,677 ± 720||0.01‡|
|Annual change in patella cartilage volume, mm3||48.5 ± 79.6||97.2 ± 91.2||0.06‡||44.4 ± 77.8||56.4 ± 59.4||0.42‡|
For the subgroup of people without patella cartilage defects at baseline (n = 192), participation in vigorous physical activity tended to be associated with a reduced risk for the development of incident patella cartilage defects (OR 0.3; 95% CI 0.1, 1.1 [P = 0.08]) (Table 5). Participation in vigorous physical activity was also associated with significantly less patella cartilage volume loss (β = −30.8; 95% CI −53.8, −7.8 [P < 0.03]), which tended to be associated with an increasing frequency (in days per week) of participation in vigorous physical activity (β = −12.7; 95% CI −27.0, 1.5 [P = 0.08]; whereby a negative regression coefficient represents a reduction in the annual loss of patella cartilage volume) (Table 5). For the subgroup of people with patella cartilage defects at baseline (n = 79), participation in vigorous physical activity was not significantly associated with patella cartilage volume change or the progression of patella cartilage defects (Table 5). These relationships persisted when people who had participated in non–weight-bearing vigorous physical activity were excluded (n = 17) (data not shown).
Table 5. Relationship between participation in vigorous physical activity (2003–2004) and incident and progression of patella cartilage defects
|No baseline patella cartilage defects (n = 192)|| || || || |
| Annual patella cartilage volume change, mm3†|| || || || |
| Participation in vigorous physical activity (yes/no)||−33.2 (−55.5, −10.9)||0.004||−30.8 (−53.8, −7.8)||0.03|
| Weekly frequency of vigorous physical activity||−15.1 (−28.9, −1.4)||0.03||−12.7 (−27.0, 1.5)||0.08|
| Incident patella cartilage defects, yes/no‡|| || || || |
| Participation in vigorous physical activity (yes/no)||0.4 (0.1, 1.3)||0.13||0.3 (0.1, 1.1)||0.08|
| Weekly frequency of vigorous physical activity||0.7 (0.3, 1.3)||0.26||0.6 (2.3, 1.2)||0.16|
|Prevalent baseline patella cartilage defects (n = 79)|| || || || |
| Annual patella cartilage volume change, mm3†|| || || || |
| Participation in vigorous physical activity (yes/no)||−7.8 (−52.0, 36.4)||0.73||−4.3 (−49.4, 40.8)||0.85|
| Weekly frequency of vigorous physical activity||−11.3 (−44.0, 21.5)||0.50||−7.6 (−41.4, 26.1)||0.65|
| Progression of patella cartilage defects, yes/no‡|| || || || |
| Participation in vigorous physical activity (yes/no)||0.6 (0.2, 2.4)||0.48||0.6 (0.1, 2.4)||0.46|
| Weekly frequency of vigorous physical activity||0.6 (0.2, 1.8)||0.32||0.5 (0.2, 1.8)||0.31|
In this longitudinal study of community-based adults with no history of knee injury or disease, participation in vigorous physical activity, which was predominantly weight-bearing in nature, was associated with a reduction in patella cartilage loss and a trend toward a reduced risk for worsening patella cartilage defects. In subgroup analyses, the benefits of participation in vigorous physical activity were apparent for people without cartilage defects at baseline, but were not observed for those with already established cartilage defects. These observations suggest that the benefits conferred by vigorous physical activity to the patellofemoral joint might be most marked for people without existing cartilage defects that signify very early joint damage. For people with baseline cartilage defects, vigorous physical activity was not significantly associated with subsequent changes to patellofemoral cartilage morphology. Larger studies examining people with and without preexisting cartilage damage are required to substantiate the findings from the subgroup analyses in this study.
The only previous study to have used MRI to examine the possible effects of physical activity on the patellofemoral compartment was cross-sectional (23). It was reported that for healthy middle-aged women, participation in fortnightly exercise sufficient to increase heart and respiratory rates for 20 minutes or more tended to be associated with a reduction in the annual rate of patella cartilage volume loss (P = 0.09) (23). At the tibiofemoral compartment, one study of delayed gadolinium-enhanced MRI of cartilage demonstrated that adults at risk of knee OA were able to increase the glycosaminoglycan content in their knee cartilage with an exercise program that was performed 3 times a week over 4 months (24). In our study, we found that participation in vigorous physical activity of a predominantly weight-bearing nature reduced the risk for worsening patella cartilage defects, as well as reduced loss of patella cartilage volume. In particular, the protective effect of vigorous physical activity on patella cartilage was observed for those with no patella cartilage defects at baseline (n = 192): participation in vigorous activity was associated with fewer incidences of patella cartilage defects and a reduction in the annual loss of patella cartilage volume. No such relationship was observed for those with prevalent patella cartilage defects at baseline, although the number of subjects was considerably smaller (n = 79). These results support participation in vigorous physical activity as being beneficial to the cartilage health of people without preexisting patella cartilage damage.
The mechanism by which vigorous physical activity of a predominantly weight-bearing nature might protect against the development of deleterious changes in patella cartilage morphology in the joint without degenerative change is unclear. Complex biomechanical factors are likely to be important since many of the physical activity tasks (e.g., running or squatting) required increased knee flexion, and thus increased patellofemoral joint load (25, 26). When the foot is in contact with the floor and the knee is flexed, a knee flexion moment is created that contributes to retropatellar load. However, as the knee is flexed, the articular contact area of the patellofemoral joint increases in an attempt to reduce the contact pressure and mitigate the retropatellar force imparted by the pull of the quadriceps and patella tendon. In the knee joint devoid of degenerative change, this antagonistic relationship between forces may ultimately create an optimal mechanical environment for articular patella cartilage. In contrast, small aberrations in either of these opposing forces may lead to cartilage degeneration. Moreover, there is evidence that mechanoreceptors located in articular cartilage might mediate aberrant chondroprotective responses in the osteoarthritic knee joint (27). It might be that in the presence of advanced cartilage damage, vigorous physical activity could be detrimental to patella cartilage, as has been simulated with in vitro animal studies, but is yet to be examined with human in vivo studies (28, 29). In any instance, there is accumulating evidence that mechanobiology is important in cartilage health (30–32). In our study, we found that the benefits conferred by participation in vigorous physical activity were apparent in the joint devoid of preexisting cartilage damage, which could point to chondroprotective responses of healthy patella cartilage to mechanical stimulation.
There are a number of potential limitations to our study. We only had a small number of individuals who demonstrated worsening patella cartilage defects over the study period, and we may not have had the power to show a relationship between physical activity and the progression of deleterious patella cartilage change in the subgroup with prevalent defects at baseline. Although we have concluded that participation in vigorous physical activity of a predominantly weight-bearing nature protects patella cartilage from deleterious change (i.e., cartilage loss and cartilage defect progression), the possibility that non–weight-bearing vigorous physical activity might have similar effects cannot be excluded, because only a very small number of people (n = 17) in this study performed non–weight-bearing vigorous physical activity. Additionally, we have not adjusted for frontal plane knee alignment, which has been shown to be associated with change in patella cartilage volume (33). Although we have performed MRI examination in the morning to mitigate any influence that diurnal variation may have on cartilage volume, we did not account for whether or not a subject had exercised vigorously in the immediate period prior to MRI, which may have temporarily influenced cartilage volume. Nonetheless, current findings suggest that human cartilage deforms very little in vivo during physiologic activities and recovers from deformation within 90 minutes after loading (34). Future studies are needed to clarify some of these issues, as well as to examine how lifelong patterns of physical activity may influence knee joint health.
A strength of our study is that questionnaires used to examine exercise of vigorous intensity have been shown to closely correlate with actual physical activity levels (35, 36). Moreover, our longitudinal measure of the entire cartilage plate (i.e., cartilage volume) circumvents reproducibility issues that may have been apparent when the dependent variable is reliant upon the reselection of a geographic location, such as when regional cartilage thickness is examined. Indeed, recent work has demonstrated that cartilage thickness and cartilage volume provide the same level of sensitivity to estimate cartilage loss in longitudinal studies, and that the potential gaining of statistical power with the use of thickness/volume change in geographic regions of the knee is negated by interpatient variability (37). However, our measures of worsening cartilage defects (yes/no) may potentially be limited since we have not necessarily monitored the progression of the same baseline defect in our longitudinal followup. Rather, we have examined a worsening defect grade for that entire compartment. However, it is anticipated that over a 2-year period, the most severe defect would have originated from a previous lesion rather than from an incident defect whose grade was severe enough to be of a greater magnitude than the baseline defect. Another strength of our study is that we excluded individuals with symptoms or a history of knee injury who might have biased our findings. On the other hand, this exclusion limits the generalizability of our results to asymptomatic people.
In this longitudinal study of community-based adults with no history of knee injury or disease, participation in vigorous physical activity, which was predominantly weight-bearing in nature, was associated with a reduced rate of patella cartilage loss and a trend toward a reduced risk for worsening patella cartilage defects. The benefits of participation in vigorous physical activity were only apparent for people without cartilage defects at baseline, and not observed for those with already established cartilage defects. This suggests that the benefits conferred by vigorous physical activity at the patellofemoral joint may be limited to people without existing cartilage defects that signify early joint damage. For people with baseline cartilage defects, vigorous physical activity was not significantly associated with subsequent changes to patellofemoral cartilage morphology. Participation in weight-bearing vigorous physical activity might therefore represent an effective preventive strategy for patellofemoral 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. Cicuttini 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. Teichtahl, Wluka, Forbes, Wang, English, Giles, Cicuttini.
Acquisition of data. Wluka, Wang, English, Giles, Cicuttini.
Analysis and interpretation of data. Teichtahl, Wluka, Forbes, Wang, Giles.
We would like to thank the study participants who made this study possible.