View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

OsteoArthritis and (2007) 15, 701e708 Crown Copyright ª 2006 Published by Elsevier Ltd on behalf of Research Society International. All rights reserved. doi:10.1016/j.joca.2006.11.004

International Cartilage Repair Society

Degenerative tears and mobility impairment in women with osteoarthritis1 A. K. Lange B.App.Sc.y*, M. A. Fiatarone Singh M.D.yzx, R. M. Smith Ph.D.y, N. Foroughi M.Sc.y, M. K. Baker B.App.Sc.y, R. Shnier M.B.B.S.k and B. Vanwanseele Ph.D.y y School of and Science, Faculty of Health Sciences, University of Sydney, Sydney, Australia z Faculty of Medicine, University of Sydney, Sydney, Australia x Hebrew SeniorLife and Jean Mayer USDA Human Nutrition Center on Aging at Tufts University, Boston, MA, USA k Symbion Clinical Research Imaging Centre, Sydney, Australia

Summary

Objective: Meniscus tears are often presumed to be associated with a traumatic event, but they can also occur as a result of the cartilage degeneration process in osteoarthritis (OA). The aim of this paper is to describe the prevalence and clinical correlates of degenerative meniscus tears in women with knee OA. Method: The subjects were women screened for a double-blind, sham-exercise controlled clinical trial for women over 40 years of age with OA in at least one knee, according to American College of Rheumatology (ACR) clinical criteria. The presence of meniscus tears was assessed via a 3 T Intera (Philips Medical Systems) magnetic resonance image (MRI). Clinical examination included a history of onset and phys- ical examination of the lower extremities. Physical assessments included body composition, muscle strength, walking endurance, gait velocity, and balance. In addition, pain and disability secondary to OA, physical self-efficacy, depressive symptoms, habitual physical activity level and quality of life were assessed via questionnaires. Results: Almost three-quarters (73%) of the 41 subjects had a medial, lateral, or bilateral by MRI. Walking endurance and balance performance were significantly impaired in subjects with a degenerative meniscus tear, compared to subjects without tears, despite similar OA duration, symptoms, and disability, body composition, and other clinical characteristics. Conclusion: Meniscus tears, diagnosed by MRI, are very common in older women with knee OA, particularly in the medial compartment. These incidentally discovered tears are associated with clinically relevant impairments of balance and walking endurance relative to subjects without meniscus tears. The explanation for this association requires further study. Crown Copyright ª 2006 Published by Elsevier Ltd on behalf of Osteoarthritis Research Society International. All rights reserved. Key words: Meniscus, Mobility, Cartilage, Osteoarthritis, Magnetic resonance imaging (MRI).

Introduction cartilage loss)5,9. Once present, meniscus tears are associ- ated with an accelerated progression of cartilage degenera- Meniscus tears are often presumed to be associated with tion in knee OA compared to individuals with OA but without a traumatic event, but they can also occur as a result of tears; presumably because of the crucial role of the menis- the cartilage degeneration process in osteoarthritis (OA) cus within the knee : distributing joint forces, load bear- of the knee. Two studies have used for menis- ing, and enhancing joint stability10,11. Thus, degenerative cus tear diagnosis and reported a prevalence of 23 and 1,2 meniscal tears are common and morbid, as they increase 53% , while two studies using magnetic resonance imag- the rate of disease progression, providing a strong rationale ing (MRI) to detect tears have reported a prevalence of 50 for a better understanding of their etiology and associated and 91%3,4. The prevalence of such tears increases with 5e8 clinical manifestations. age and the severity of OA (defined by the degree of The focus of previous research has been on the radio- graphic and clinical exam presentation of degenerative menis- 1Grant support: University of Sydney: Cumberland Grant e Category cus tears. To our knowledge, comprehensive comparisons of B to Vanwanseele, B, January 2005; R&D Grant (S4201 U3301) to the physiological profile and health status of arthritic individ- Vanwanseele, B, January 2006. uals with and without degenerative meniscus tears have not *Address correspondence and reprint requests to: Angela Lange, been published. Body composition, fitness, and physical/ Exercise and Sport Science, Faculty of Health Science, University of Sydney, PO Box 170, Lidcombe, NSW 1825, Australia. Tel: 612- psychological function may potentially be related to the occur- 9351-9279; Fax: 612-9351-9204; E-mail: [email protected]. rence of meniscal tears, as these factors are related to the edu.au etiology, clinical and pathophysiological severity of OA itself. Received 13 September 2006; revision accepted 18 November For example, the direct relationship between body weight or 2006. body mass index (BMI) and knee OA is well established in

701 702 A. K. Lange et al.: Meniscus tears and mobility women12; quadriceps weakness is associated with abnormal time ¼ 35 ms, flip angle ¼ 90, slice thickness ¼ 2.2 mm, in- joint loading and instability of the knee, which can lead to the plane resolution ¼ 0.13 mm) (normal meniscal signal is development and progression of knee OA13; aerobic capacity black) originating within the meniscus and extending through and habitual physical activity levels tend to decline due to knee one or both of its articular surfaces of the meniscus (see pain14; proprioception required for balance is decreased in Fig. 1). A second type of tear was detected when a piece of individuals who suffer from degenerative joint diseases15, the meniscus had ‘‘fallen off’’ and was displaced elsewhere and depressive symptoms and low self-efficacy are associ- in the joint. The type (e.g., radial) and location (e.g., posterior ated with pain and disability in OA16,17. It is important to iden- horn) of each tear were read, but due to the sample size stud- tify potentially remediable etiological factors in this cohort to ied, only the presence and the compartment in which the tear develop effective preventive strategies, as well as to identify was located were included in analyses. Images were exam- potential clinical consequences of such tears, so that these ined by one highly experienced radiologist (RS) and indepen- factors may be considered in comprehensive management dently confirmed by BV; both were blinded to all details and rehabilitative strategies. except subjects’ age. We hypothesized that individuals with OA of the knee Clinical evaluation included a history of arthritis onset and associated meniscus tears would be older, have and a complete physical examination of the lower a higher BMI and percent body fat, lower fat-free mass, extremities. Blinded physician assessors conducted reduced exercise capacity (muscle weakness, reduced clinical evaluation and MRI interpretation independently walking endurance, and impaired balance), lower habitual of each other. All MRI and clinical data reported were physical activity levels poorer health-related quality of life, obtained at screening prior to any exercise testing/training a greater burden of chronic disease and depressive symp- procedures. toms, and higher disability and pain rating scores, com- pared to individuals with OA of the knee but without a meniscus tear. Therefore, we conducted a cross-sec- INDEPENDENT VARIABLES tional analysis of a cohort of women undergoing screening Dynamic muscle strength for possible inclusion in a randomized controlled trial of ex- ercise treatment for OA of the knee to investigate these Lower extremity peak strength was assessed using hypotheses. digital K400 Keiser pneumatic resistance machines (Keiser Health Equipments, Inc., Fresno, CA). One repeti- tion maximum (1RM) tests were performed according to de Vos et al.19 unilaterally on knee extension, abduction Patients and methods and adduction; and bilaterally on knee flexion, leg press, STUDY DESIGN and plantar flexion. Strength tests were performed twice at baseline approximately 1 week apart, and the higher of Cross-sectional study of baseline data from subjects the two results was recorded as the 1RM. The mean (range) screened for a randomized controlled trial. coefficient of variation (CV) for the nine in this cohort was 13.4 (9.8e21.7)%. STUDY POPULATION The subjects were volunteering for a double-blind, sham-exercise controlled clinical trial for community-dwell- ing women over 40 years of age in stable health with pri- mary OA in at least one knee, according to the American College of Rheumatology (ACR) clinical criteria18. Partici- pants were excluded if they had secondary OA (i.e., OA diagnosed due to trauma, surgery, or other disease pro- cess); joint , injection or surgery within the past 6 months or ; already participated in structured exercise for more than 1 day/week during the previous 3 months; any contraindications to exercise and/or MRI; severe functional limitation (unable to walk without the assistance of a person); or cognitive impairment. The University of Sydney Human Research Ethics Com- mittee approved testing procedures on 13 December 2004 (Reference No. 12-2004/2/7848) and written informed con- sent was obtained from all subjects. This study has been lodged with the Australian Clinical Trials Registry (ACTR Reference No. 12605000116628).

PRIMARY DEPENDENT VARIABLE: MENISCUS PATHOLOGY Meniscus pathology in the knee with the most severe clin- ical signs and symptoms was assessed via a 3 T Intera MR (Philips Medical Systems, Achieva 3T) image in the coronal plane. Meniscus tears were diagnosed when an abnormally bright signal was seen either on the three-dimensional Fig. 1. Coronal view of the tibiofemoral joint. Arrow is pointing to T2-weighted sequence (repetition time ¼ 4418 ms, echo a tear. Osteoarthritis and Cartilage Vol. 15, No. 6 703

Walking endurance index indicated better balance (less sway, longer stance time): The 6-min walk distance was used to assess walking endurance. The 6-min walk was performed according to Balance Index ¼ Sum of 12 sway measures þ Guyatt et al.20. All participants were given specific instruc- ð180 Sum of 6 time measuresÞ tions e ‘‘To cover as much ground as possible by walking as quickly as you can’’ with verbal encouragement every e 30 s. The better of two trials 1 week apart was recorded. This index has been shown to have high test retest reli- The CV in this cohort was 3.0 (0.0e13.0)%. ability (correlation of repeated assessments 10 weeks apart in older adults with stable chronic disease (r ¼ 0.759, P 0.001)) and is sensitive to change with an exercise Habitual gait velocity intervention24. Gait analysis was performed using a 10-camera Motion Analysis system (Santa Rosa, California) set to sample at Questionnaires 100 Hz. Thirty-eight passive reflective markers were placed bilaterally on standard bony landmarks of the lower All questionnaires were interviewer-administered in a and upper body. Subjects were asked to walk at their self- private room using visual prompts. selected normal speed and five trials were collected for Symptomatology and disability was assessed using the each subject. Gait velocity was defined as the mean hor- Western Ontario and McMaster Osteoarthritis Index 25 izontal velocity of the sacrum marker during two full strides (WOMAC) questionnaire . Habitual physical activity levels and was averaged over five trials. The CV in this cohort were assessed using the Physical Activity Scale for the El- 26 was 4.6 (0.7e12.5)%. derly (PASE) questionnaire and Harvard Alumni Question- naire27. Depressive symptoms were assessed using the Geriatric Depression Scale (GDS)28. Physical self-efficacy Body composition was assessed for lifting objects, walking, jogging, climbing stairs, and doing push-ups, all at a steady pace without BMI was calculated from fasting weight and stretched 29 height measurements21. Waist circumference was pausing/stopping using the Ewart Self-Efficacy Scale . measured according to the International Society for the Health-related quality of life was assessed using Version 2 21 of the Medical Outcome Survey 36-item Short-Form (SF- Advancement of Kinathropometry (ISAK) protocol . Per- 30 cent body fat and fat-free mass were estimated using 36) . The SF-36 produced eight health outcome measures: bioelectrical impedance (BIA-101: RJL Systems, Detroit, three physical parameters [physical functioning (PF), limita- MI). All participants were measured three times early in tions due to physical difficulties (RP), and bodily pain (BP)]; the morning after a 12-h fast. The CV in this cohort for resis- three mental health parameters [mental health (MH), role tance was 0.03 (0.0e0.1)%. Fat mass and fat-free mass limitations due to emotional difficulties (RE), and social func- were calculated from the formula developed by Lukaski tioning (SF)]; and two combined physical and mental health et al. for older adults22. parameters [general health (GH) and vitality (VT)].

Balance STATISTICAL ANALYSIS The Chattecx Dynamic Balance System (software Data were inspected for normality visually and statistically version 4.20; Chattecx Corp, Chattanooga Group Inc., (skewness 1 1), and expressed as mean and standard Hixson, TN) was used to assess balance. This system deviation or median and range, as appropriate. Nonnor- allows testing of static balance time and body sway via mally distributed data were log-transformed prior to use measurements from the force platform23. Three test condi- with parametric statistics if possible or used with nonpara- tions were performed with eyes open and with eyes closed metric tests if assumptions of normality were not met de- in a random order: (1) narrow bilateral stance on the plat- spite transformation. Comparisons between groups were form sliding backward and forward at a speed of 8.3 made using Chi-Square tests for categorical data and anal- s/cycle in the anterioreposterior direction; (2) narrow bilat- ysis of variance (ANOVA) for normally distributed continu- eral stance on the platform tilting up and down from 0 to ous data. The KruskaleWallis test or ManneWhitney 2 in the anterioreposterior direction; and (3) unilateral U (MWU) test were used for nonnormally distributed contin- stance of the preferred leg on still platform. Balance was uous data. Pairwise comparisons of means from analysis of tested for up to 30 s/test. A maximum of three trials covariance (ANCOVA) models with significant F-ratios were were allowed to complete each test if participants lost their conducted using two-tailed least significant difference balance (touching hand rails, taking a step off the platform, (LSD) post-hoc t tests. A post-hoc comparison of medians requiring support from the assessor). If no attempt was from significant KruskaleWallis tests was carried out by successful, the trial with the longest time was recorded sequentially applying the KruskaleWallis test to all pairs. and only data from this trial were analyzed. Unilateral SPSS (Release 13.0 for Windows, 2004, Chicago: SPSS stance duration with eyes closed, summated maximum Inc) was used for all data analysis; except for SF-36 data sway in four directions (medial, lateral, anterior and poste- for which StatView was used (Version 5.0 for Windows, rior), and number of trials needed to complete the six Cary, NC: SAS Institute Inc). All P values of less than conditions were used as individual measures of balance 0.05 were considered statistically significant except for performance. post-hoc comparisons of pairs of variables from significant Overall balance performance was examined using KruskaleWallis models, which were adjusted for multiple a balance index24. The balance index was calculated by comparisons using the method of Bonferroni. Clinical mean- summating all anterioreposterior and medio-lateral sway ingfulness of differences observed was assessed by evalu- measures and time results. In order to remain consistent ation of the magnitude of the differences relative to clinical with the direction of sway values, a lower balance outcomes in the literature. 704 A. K. Lange et al.: Meniscus tears and mobility

Results Additionally, mobility (walking endurance and balance) was significantly worse in subjects with meniscus tears, RECRUITMENT as detailed below. However, no other measured clinical Flow of participants is shown in Fig. 2. Forty-three characteristics (including OA duration, symptoms of pain percent of those screened had an MRI and clinical exam and disability, laxity, burden of other chronic dis- and were included in the analyses, and 32% completed eases, strength, psychological health, body composition, all baseline assessments. or level of obesity) were associated with the presence of a meniscus tear.

HEALTH STATUS AND PHYSICAL FUNCTION WALKING ENDURANCE Baseline characteristics are presented in Table I. The co- hort was primarily overweight or obese (84%), and had suf- Participants without a meniscus tear walked over 55 m fered from OA for 10 years on average. Average measures further than participants with a tear (F ¼ 4.515, P ¼ 0.04) of health-related quality of life were within the normal range during the 6-min walk test. When adjusted for age, the sta- of 50 1030; however, 46% of subjects were outside this tistical significance of this relationship was slightly attenu- range (1 standard deviation (SD) below normal) for phys- ated (F ¼ 3.663, P ¼ 0.063). Similarly, when analyzed with ical function difficulties and bodily pain categories. Less respect to the location of the tear, subjects without tears than 10% of the cohort had depressive symptom scores walked almost 100 m further than subjects with bilateral in the depressed range. Average gait velocity was below tears (560.9 65.6 m vs 469.9 90.2 m; F ¼ 2.566, the normal range identified for older adults (1.18e1.45 P ¼ 0.069). m/s)31. There was a large range in lower extremity strength By contrast, habitual gait velocity over two strides was and habitual levels of physical activity across the cohort. similar between those with and without a tear (1.08 vs Notably, habitual energy expenditure in physical activity 1.10 m/s), both unadjusted (F ¼ 0.081, P ¼ 0.778) and was very low, approximately 50% of the desired level adjusted (F ¼ 0.024, P ¼ 0.879) for age. (2000 kcal/week) associated with decreased mortality in adults32; and 42% of subjects were below 500 kcal/week, BALANCE placing them within the highest risk category for cardiovas- cular disease and total mortality32. In general, subjects without tears performed better in all Using linear regression analysis, older age was associ- measures of balance, as detailed below. Overall balance ated with poorer balance performance (Balance Index) performance (Balance Index) was also significantly better (r ¼ 0.438, P ¼ 0.004), and shorter 6-min walk distance (i.e., less anterioreposterior and medio-lateral sway, longer was associated with lower physical activity levels measured stance duration) in participants without tears compared to via the Harvard Alumni Questionnaire (r ¼ 0.337, those with a tear (F ¼ 4.514, P ¼ 0.040). When adjusted P ¼ 0.031). for age, the relationship between balance and the presence of a tear was even more statistically significant (F ¼ 8.565, P ¼ 0.006). PREVALENCE OF MENISCAL TEARS AND CLINICAL As shown in Fig. 3 there was also an effect of tear loca- CORRELATES tion on balance (F ¼ 3.392, P ¼ 0.028). After adjustment for As shown in Table II, almost three-quarters (73%) of par- age this relationship remained statistically significant ticipants screened had a medial (53%), lateral (20%) or bi- (F ¼ 3.474, P ¼ 0.026). The lateral tear group had the poor- lateral (27%) meniscus tear. Eighty-five percent of tears est balance, significantly worse than no tear (P ¼ 0.004) were located in the same compartment as the OA (defined and bilateral tear groups (P ¼ 0.033), and tended to be by cartilage degeneration and presence of osteophytes) worse than the medial tear group (P ¼ 0.066). was located. Subjects with tears were significantly (approx- There was also a trend for unilateral static balance time to imately 5 years) older than subjects without tears. be better in those without a tear (median (range), 3.53

Telephone Screened Ineligible n= 129 n= 58

Physician Screened n= 65

Ineligible Unable to meet study Ineligible criteria Completed MRI Did not meet ACR n=7 n= 54 criteria Medical= 4 n=3 Other=3

Completed Baseline Assessments n= 41

Fig. 2. Participant flow through the study. Osteoarthritis and Cartilage Vol. 15, No. 6 705

(1e21.46) s) compared to subjects with a tear (2.77 (1.25e7.53) s; MWU ¼ 106.0, P ¼ 0.083). Table I Finally, medio-lateral sway amplitude tended to be better Baseline characteristics (lower) in subjects without tears (P ¼ 0.107). Characteristic Mean SD Age (years) 63 9 BMI (kg/m2)* 33.1 6.8 Discussion Normal (%) 18.5e24.99 15.8 Overweight (%) 25.00 15.8 We have observed that meniscus tears, diagnosed inci- Obese Class I (%) 30.00e34.99 42.1 dentally by MRI, are very common in older women with Obese Class II (%) 35.00e39.99 13.2 knee OA, particularly in the medial compartment, as has Obese Class III (%) 40 13.2 been shown by others6,9. None of the tears in our cohort Body fat (%) 44.1 7.5 were clinically diagnosed or suspected by referring physi- Fat-free mass (kg) 46.6 6.1 Waist circumference (cm) 98.8 14.8 cians prior to our study MRI. Subjects with tears were older, Duration of OA (yrs)y 10 (1e31) and had impaired walking endurance and balance com- No. chronic diseases (n) 1.8 1.5 pared to those with a similar severity of OA (i.e., similar du- WOMAC ration, pain, stiffness, and associated disability), but no e Pain (range 0 20) 6 3 tears. This is to our knowledge the first time clinically unsus- Stiffness (range 0e8) 4 2 Difficulty (range 0e68) 23 11 pected degenerative meniscus tears associated with OA Total (range 0e96) 33 15 have been shown to be related to mobility impairment. GDSy (range 0e30) 2 (0e22) The magnitude of the differences in mobility between the in- Ewart self-efficacy (0e100%) dividuals with and without tears are clinically relevant. The Lifting objects (%) 42 23 differences in balance performance between these two Walking (%) 59 24 groups in our study are even greater than those seen in Jogging (%) 20 24 a cohort of elderly individuals following a 10-week power Climbing stairs (%) 43 27 24 Push-ups (%) 14 16 training program . The difference in walking endurance be- SF-36 (0e100) tween those with and without a tear was approximately Physical functioning 41 9 11%; this is just below the average 15% change in walking Role limitations 48 6 endurance seen following participation in a short-term car- due to physical difficulties diac rehabilitation program33. Bodily pain 43 6 General health 46 8 Even though the meniscus tears appeared initially Vitality 47 9 asymptomatic and clinically ‘‘silent’’ (discovered incidentally Social functioning 50 8 on MRI), they were in fact not clinically silent, as they were Role limitations 52 5 associated with clinically relevant mobility impairment when due to emotional difficulties detailed research testing was performed in addition to stan- Mental health 54 9 Habitual physical activity 135.6 56.4 dard clinical history and physical examination. This high- level (PASE) lights the need for physicians to be aware of not only the Harvard Alumni Questionnaire 1044.2 1192.6 ‘‘typical’’ clinical presentation of a degenerative meniscus (kcal/week) tear, but also the potential mobility impairment associated Strength with such tears. This is particularly relevant for the preven- Knee extension (N m) Left 38 19 tion of ; as poor balance performance increases the Right 38 18 34 Knee flexion (N m) Bilateral 93 26 risk of falls/fractures in the elderly . Leg press (N) Bilateral 523 171 Although advancing age has been proposed as a risk fac- 8 Plantar flexion (N) Bilateral 488 139 tor for meniscal tears due to cartilage degeneration , lower Hip abduction (N) Left 87 26 walking endurance and impaired balance performance in in- Right 78 20 dividuals with meniscus tears may be either a risk factor or Hip adduction (N) Left 105 25 a consequence of such tears. Evidence to support any the- Right 106 35 6-min walk (m) 519.5 79.0 ory on the etiology of meniscus tears is beyond the scope of Gait velocity (m/s) 1.09 0.17 this cross-sectional study, however, we propose two poten- Balance index 91.9 19.7 tial pathways [Fig. 4(A,B)] as possible reasons for the Single-leg stance duration: 3.06 (1e21.46) greater degree of mobility impairment observed in our co- eyes closed (s)y hort with tears. No. balance trials (n) 9.7 1.6 (range 6e18) In the first model [Fig. 4(A)], the combination of OA and age leads to a reduction in physical activity participation n ¼ 41, except for age, duration of OA and number of chronic dis- and impairments in mobility. The vicious cycle that exists eases where n ¼ 54. between reduced activity levels and overall mobility impair- WOMAC ¼ the lower the score the less symptomatic and physi- ment contributes to the progression of OA and ultimately the cally impaired. development of a degenerative meniscus tear. GDS ¼ 0e9normal,>9e30 depressive range. In the second model [Fig. 4(B)], age and the development Ewart Self-Efficacy Scale; scale ranges from 0 (not at all confident) and progression of OA contribute to the development of to 100 (completely confident) with intervals of 10. a degenerative meniscus tear. The excess pain and disabil- Nm¼ Newton meters. N ¼ Newtons. ity associated with such tears limit individuals’ ability to par- *Adapted from World Health Organization (WHO) 1995, WHO 2000 ticipate in physical activity more than individuals with OA and WHO 2004. but no tears, and the subsequent deconditioning results in yReported as median (range) data not normally distributed. accelerated mobility impairment. In both proposed models, increased age is a risk factor for degenerative meniscus tears. This is supported by our 706 A. K. Lange et al.: Meniscus tears and mobility

Table II Characteristics of subjects with and without degenerative meniscus tears Characteristic (N ¼ 41) Mean SD P value Tear No tear Meniscus tear (%) 73 27 Age (years) 65 860 10 0.037 BMI (kg/m2)* 33.3 6.9 32.4 6.8 0.693 Fat-free mass (kg) 46.0 5.3 48.1 7.8 0.346 Duration of OA (years)y 10.1 6.9 10.9 6.7 0.749 No. chronic diseases (n) 1.7 1.4 1.4 1.3 0.495 WOMAC Pain (range 0e20) 6.5 3.2 5.3 2.6 0.266 Stiffness (range 0e8) 3.8 1.5 3.6 1.6 0.713 Difficulty (range 0e68) 23.9 10.6 21.1 11.7 0.474 Total (range 0e96) 34.2 14.5 29.9 15.2 0.422 Habitual physical activity level (PASE) 136.0 56.0 134.4 60.5 0.935 Harvard Alumni 929.9 1117.7 1356.0 1398.1 0.318 Questionnaire (kcal/week) Knee extension Left 37 16 40 27 0.681 Right 36 14 43 25 0.335 6-min walk (m) 504.2 79.0 560.9 65.6 0.040 Gait velocity (m/s) 1.08 0.16 1.10 0.21 0.778 Balance index 95.6 18.8 81.5 19.1 0.040 No. balance trials (n) (range 6e18) 10.0 1.7 8.8 1.07 0.041

*Adapted from World Health Organization (WHO) 1995, WHO 2000 and WHO 2004. yReported as median (range) data not normally distributed.

e study and many others in the literature5 8. The primary research, has shown that pain, stiffness and self-reported distinguishing feature between the proposed models is the limitations in physical function, assessed via the WOMAC timing of worsening mobility impairment relative to the de- questionnaire, were unrelated to the presence of tears48. velopment of the degenerative meniscus tear. Therefore, overt symptoms of a degenerative meniscus In Model 1, excessive impairments in mobility are risk fac- tear are unlikely to cause the reduction in physical activity tors for a degenerative meniscus tear. The evidence from and subsequent impairments in mobility seen in subjects our study and previous literature favor Model 15,9,13,35e47; with meniscus tears. and are consistent with the causal pathway between mobil- ity impairment and degenerative meniscus tears. In contrast, Model 2 suggests that impairments in mobility A are a consequence of the excess symptoms (pain and sub- Model 1 sequent activity restriction) attributable to the development OA+Age of a degenerative meniscus tear. Model 2 is unlikely to be the reason for the mobility impairment seen in our cohort Activity primarily because our study, consistent with previous Cartilage Degeneration Mobility (6MW, Balance)

Tear *** ** B Model 2 140 * OA+Age 120 No Tear 100 Cartilage Degeneration 80 Bilateral 60 Tear Medial 40 Balance Index 20 Lateral Pain/Disability/Abnormal Gait Mechanics 0 Tear Location Activity Fig. 3. Balance index in subjects with and without meniscus tears. A lower score for the balance index indicates better overall balance performance (less anterioreposterior and medio-lateral sway, Mobility (6MW, Balance) longer stance duration). ANOVA model: F ¼ 3.392, P ¼ 0.028 (unadjusted for age); F ¼ 3.474, P ¼ 0.026 (age-adjusted). LSD post-hoc t tests were used for all pairwise comparisons of no Fig. 4. (A,B) Theoretical models of the potential relationship be- tear, medial, lateral, and bilateral meniscus tears. *P ¼ 0.066, tween impairments in mobility and degenerative meniscus tears. **P ¼ 0.033, and ***P ¼ 0.004. 6MW ¼ 6-min walk. Osteoarthritis and Cartilage Vol. 15, No. 6 707

It is possible that pathology in other joint structures linked 9. Noble J, Hamblen DL. The pathology of the degenerate to cartilage tears was the actual cause of the balance and meniscus lesion. J Bone Joint Surg Br 1975;57(2): walking impairment. For example, ligament disruption or lax- 180e6. ity would potentially predispose to both balance impairment 10. Biswal S, Hastie T, Andriacchi TP, Bergman GA, and cartilage tears49,50. We found no disruption of ligament Dillingham MF, Lang P. Risk factors for progressive integrity in any subjects by MRI scan, as all anterior and pos- cartilage loss in the knee: a longitudinal magnetic res- terior cruciate and medial and lateral collateral onance imaging study in forty-three patients. Arthritis were intact. Ligament laxity was also assessed qualitatively Rheum 2002;46(11):2884e92. during the physical examination by the study physician, and 11. Hunter DJ, Zhang YQ, Niu JB, Tu X, Amin S, Clancy M, was noted in three subjects, all of whom had tears. However, et al. The association of meniscal pathologic changes future research is needed to explore potential relationships with cartilage loss in symptomatic knee osteoarthritis. between ligament integrity, other structures in and around Arthritis Rheum 2006;54(3):795e801. the knee such as muscle strength, and the presence 12. Felson DT. Risk factors for osteoarthritis. Clin Orthop of meniscus tears and mobility impairment. 2004;427(Suppl):S16e21. In conclusion, this cross-sectional study suggests that 13. Baker KR, Lea Xu. Quadriceps weakness and its rela- degenerative meniscus tears measured by MRI are strongly tionship to tibiofemoral and patellofemoral knee osteo- associated with clinically relevant mobility impairment in arthritis in Chinese: the Beijing osteoarthritis study. women with knee OA, relative to OA subjects without tears. Arthritis Rheum 2004;5(6):1815e21. Future longitudinal research into the mechanisms by which 14. Ettinger WHJ, Afable RF. Physical disability from knee reduced activity levels and altered mobility and balance osteoarthritis: the role of exercise as an intervention. may predispose to the development of degenerative menis- Med Sci Sports Exerc 1994;26(12):1435e40. cus tears requires further investigation. 15. Jerosch J, Prymka M. Proprioception and joint stability. Knee Surg Sports Traumatol Arthrosc 1996; 4:171e9. Acknowledgments 16. Husaini BA, Moore ST. Arthritis disability, depression, and life satisfaction among black elderly people. The authors would like to acknowledge Keiser Sports Health Soc Work 1990;15(4):253e9. Health, Inc (Fresno, CA, USA) for the donation of the 17. Rejeski WJ, Craven T, Ettinger WHJ, McFarlane M, K400 Electronics for the strength training equipment; the Shumaker S. Self-efficacy and pain in disability with STRONG Clinic at Balmain Hospital for their assistance osteoarthritis of the knee. J Gerontol B Psychol Sci with training personnel; and the Symbion Clinical Research 1996;51B(1):24e9. Imaging Centre for their help with processing the MRIs. 18. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Finally, the authors would like to thank the wonderful Brandt K, et al. The American College of Rheumatol- volunteers who have devoted their time to this project. ogy criteria for the classification and reporting of oste- oarthritis of the knee. Arthritis Rheum 1986;29: e References 1039 49. 19. de Vos NJ, Singh NA, Ross DA, Stavrinos TM, Orr R, 1. Bohnsack M, Lipka W, Ruhmann O, Peters G, Fiatarone Singh MA. Optimal load for increasing mus- Schmolke S, Wirth C. The value of knee arthroscopy cle power during explosive resistance training in older in patients with severe radiological osteoarthritis. adults. J Gerontol A Biol Sci Med Sci. 2005;60A: Arch Orthop Trauma Surg 2002;122(8):451e3. 638e47. 2. Timoney JM, Kneisl JS, Barrack MC, Alexander AH. 20. Guyatt GH, Sullivan MJ, Thompson PJ, Fallen EL, Arthroscopy in the osteoarthritic knee: long-term fol- Pugsley SO, Taylor DW, et al. The 6-minute walk: low-up. Orthop Rev 1990;19(4):371e9. a new measure of exercise capacity in patients with 3. Ali G, Souhil Z, Bachir T, Yves M, Charles GP, chronic heart failure. CMAJ 1985;132:919e23. Harry KG. MR findings in knee osteoarthritis. Eur 21. Marfell-Jones M, Olds T, Stewart A, Carter JEL. Inter- Radiol 2003;V13(6):1370e86. national Standards for Anthropometric Assessment. 4. Bhattacharyya T, Gale D, Dewire P, Totterman S, The International Society for the Advancement of Gale ME, McLaughlin S, et al. The clinical importance Kinanthropometry 2006. of meniscal tears demonstrated by magnetic reso- 22. Lukaski HC, Bolunchuk WW, Hall CB, Siders WA. Val- nance imaging in osteoarthritis of the knee. J Bone idation of tetrapolar bioelectrical impedance method to Joint Surg Am 2003;85(1):4e9. assess human body composition. J Appl Physiol 1986; 5. Christoforakis J, Pradhan R, Sanchez-Ballester J, 60:1327e32. Hunt N, Strachan RK. Is there an association between 23. Mattacola CG, Lebsack DA, Perrin DH. Intertester articular cartilage changes and degenerative menis- reliability of assessing postural sway using the Chat- cus tears? Arthroscopy 2005;21(11):1366e9. tecx Balance System. J Athl Train 1995;30(3): 6. Fukuta S, Masaki K, Korai F. Prevalence of abnormal 237e9. 41e42. findings in magnetic resonance images of asymptom- 24. Orr R, De Vos NJ, Singh NA, Ross DA, Stavrinos TM, atic . J Orthop Sci 2002;7(3):287e91. Fiatarone Singh MA. Power training improves balance 7. Jerosch J, Castro WH, Assheuer J. Age-related mag- in healthy older adults. J Gerontol A Biol Sci Med Sci netic resonance imaging morphology of the menisci 2006;61A(1):78e85. in asymptomatic individuals. Arch Orthop Trauma 25. Bellamy N, Buchanan W, Goldsmith CH, Campbell J, Surg 1996;115(3e4):199e202. Stitt LW. Validation study of WOMAC: a health status 8. Kornick J, Trefelner E, McCarthy S, Lange R, Lynch K, instrument for measuring clinically important patient Jokl P. Meniscal abnormalities in the asymptomatic relevant outcomes to antirheumatic drug therapy in population at MR imaging. Radiology 1990;177(2): patients with osteoarthritis of the hip or the knee. 463e5. J Rheumatol 1988;15(12):1833e40. 708 A. K. Lange et al.: Meniscus tears and mobility

26. Washburn R, Smith K, Jette A. The physical activity 40. Isles RC, Low Choy NL, Steer M, Nitz JC. Normal scale for the elderly (PASE): development and evalu- values of balance tests in women aged 20e80. J Am ation. J Clin Epidemiol 1993;48:153e62. Geriatr Soc 2004;52:1367e72. 27. Paffenbarger RS, Wing AL, Hyde RT. Physical activity 41. Lord SR, Menz HB. Physiologic, psychologic, and as an index of heart attack risk in college alumni. health predictors of 6-minute walk performance in Am J Epidemiol 1978;108:161e75. older people. Arch Phys Med Rehabil 2002;83(7): 28. Brink TL, Yesavage JA, Lum O, Heersema P, Adey MB, 907e11. Rose TL. Screening tests for geriatric depression. Clin 42. Messier SP, Loeser RF, Miller GD, Morgan TM, Gerontol 1982;1:37e44. Rejeski WJ, Sevick MA, et al. Exercise and dietary 29. Ewart C, Stewart K, Gillian R, Keleman M. Self efficacy weight loss in overweight and obese older adults mediates strength gains during circuit weight training with knee osteoarthritis: the arthritis, diet, and activity in men with coronary artery disease. Med Sci Sports promotion trial. Arthritis Rheum 2004;50(5):1501e10. Exerc 1986;18(15):531e40. 43. Samson M, Meeuwsen I, Crowe A, Dessens J, 30. Ware JE, Kosinski M, Dewey JE. How to Score Version Duursma S, Verhaar H. Relationships between physi- 2 of the SF-36 Health Survey. Lincoln, RI: QualityMet- cal performance measures, age, height and body ric Incorporated 2000. weight in healthy adults. Age Ageing 2000;29(3): 31. Daley MJ, Spinks WL. Exercise, mobility and aging. 235e42. Sports Med 2000;29(1):1e12. 44. Shilke JM, Johnson GO, Housh TJ, O’Dell JR. Effects 32. Paffenbarger RS, Hyde RT, Wing AL, Hsieh CC. Phys- of muscle-strength training on the functional status of ical activity, all-cause mortality, and longevity of col- patients with osteoarthritis of the knee joint. Nurs lege alumni. N Engl J Med 1986;314(10):605e13. Res 1996;45(2):68e72. 33. Verrill DE, Barton C, Beasley W, Lippard M, King CN. 45. Vanwanseele B, Eckstein F, Knecht H, Spaepen A, Six-minute walk performance and quality of life com- Stu¨ssi E. Longitudinal analysis of cartilage atrophy in parisons in North Carolina cardiac rehabilitation pro- the knees of patients with spinal cord injury. Arthritis grams. Heart Lung 2003;32(1):41e51. Rheum 2003;48(12):3377e81. 34. Tinetti ME, Speechley M, Ginter SF. Risk factors for 46. Vanwanseele B, Lucchinetti E, Stu¨ssi E. The effects of falls among elderly persons living in the community. immobilization on the characteristics of articular carti- N Engl J Med 1988;319(26):1701e7. lage: current concepts and future directions. Osteoar- 35. Buchner DM. Effects of physical activity on health thritis Cartilage 2002;10(5):408e19. status in older adults: intervention studies. Annu Rev 47. Wolfson L. Gait and balance dysfunction: a model of Publ Health 1992;13:439e88. the interaction of age and disease. Neuroscientist 36. Camicioli R, Panzer VP, Kaye J. Balance in the healthy 2001;7(2):178e83. elderly: posturography and clinical assessment. Arch 48. Kornaat PR, Bloem JL, Ceulemans RY, Riyazi N, Neurol 1997;54(8):976e81. Rosendaal FR, Nelissen RG, et al. Osteoarthritis of 37. Fontaine KR, Haaz S. Risk factors for lack of recent the knee: association between clinical features and exercise in adults with self-reported, professionally MR imaging findings. Radiology 2006;239(3):811e7. diagnosed arthritis. J Clin Rheumatol 2006;12(2): 49. Ageberg E, Roberts D, Holmstrom E, Friden T. Balance 66e9. in single-limb stance in patients with anterior cruciate 38. Hirata S, Ono R, Yamada M, Takikawa S, Nishiyama T, ligament injury: relation to knee laxity, proprioception, Hasuda K, et al. Ambulatory physical activity, disease muscle strength, and subjective function. Am J Sports severity, and employment status in adult women with Med 2005;33(10):1527e35. osteoarthritis of the hip. J Rheumatol 2006;33(5): 50. Cipolla M, Scala A, Gianni E, Puddu G. Different patterns 939e45. of meniscal tears in acute anterior cruciate ligament 39. Hunter GR, McCarthy JP, Bamman MM. Effects of re- (ACL) ruptures and in chronic ACL-deficient knees. sistance training on older adults. Sports Med 2005; Classification, staging and timing of treatment. Knee 34(5):329e48. Surg Sports Traumatol Arthrosc 1995;3(3):130e4.