Advertisement

Joint space narrowing, body mass index, and knee pain: the ROAD study (OAC1839R1)

Open ArchivePublished:January 29, 2015DOI:https://doi.org/10.1016/j.joca.2015.01.011

      Summary

      Objective

      The objective of the present study was to clarify the association of joint space narrowing with knee pain in Japanese men and women using a large-scale population-based cohort of the Research on Osteoarthritis/osteoporosis Against Disability (ROAD) study.

      Methods

      This study examined the association between minimum joint space width (mJSW) in the medial compartment and pain at the knee. mJSW was measured in the medial and lateral compartments of the knee using a knee osteoarthritis (OA) computer-aided diagnosis system.

      Results

      From the 3040 participants in the ROAD study, the present study analyzed 2733 participants who completed the radiographic examinations and questionnaires regarding knee pain (975 men and 1758 women; mean age, 69.9 ± 11.2 years). Subjects with lateral knee OA were excluded. After adjustment for age and Body mass index (BMI), medial mJSW, as well as medial mJSW/lateral mJSW, was significantly associated with knee pain. Sex and BMI affected the association of medial mJSW with knee pain. The threshold of medial mJSW was approximately 3 mm in men and 2 mm in women, while that of medial mJSW/lateral mJSW was approximately 60% in both men and women. BMI was found to have a distinct effect on the association of mJSW with pain.

      Conclusion

      The present cross-sectional study using a large-scale population from the ROAD study showed that joint space narrowing had a significant association with knee pain. The thresholds of joint space narrowing for knee pain were also established.

      Keywords

      Introduction

      Knee osteoarthritis (OA) is a major public health issue that causes chronic pain and disability
      • Sharma L.
      • Kapoor D.
      Epidemiology of osteoarthritis.
      • Guccione A.A.
      • Felson D.T.
      • Anderson J.J.
      • Anthony J.M.
      • Zhang Y.
      • Wilson P.W.
      • et al.
      The effects of specific medical conditions on the functional limitations of elders in the Framingham Study.
      • Felson D.T.
      • Zhang Y.
      An update on the epidemiology of knee and hip osteoarthritis with a view to prevention.
      . The prevalence of radiographic knee OA is high in Japan
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      , with 25,300,000 persons aged 40 years and older estimated to have radiographic knee OA
      • Yoshimura N.
      • Muraki S.
      • Oka H.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of knee osteoarthritis, lumbar spondylosis and osteoporosis in Japanese men and women: the research on Osteoarthritis/Osteoporosis Against Disability (ROAD).
      . According to the recent National Livelihood Survey of the Ministry of Health, Labour and Welfare in Japan, OA is ranked fourth among diseases that cause disabilities that subsequently require support with activities of daily living
      • Ministry of Health, Labour and Welfare
      The Outline of the Results of National Livelihood Survey.
      .
      Knee pain is the principal clinical symptom of knee OA
      • Linaker C.H.
      • Walker-Bone K.
      • Palmer K.
      • Cooper C.
      Frequency and impact of regional musculoskeletal disorders.
      . Although much effort has been devoted toward a definition of knee pain, the correlation with radiographic severity of the knee OA was not as strong as one would expect
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      • Summers M.N.
      • Haley W.E.
      • Reveille J.D.
      • Alarcon G.S.
      Radiographic assessment and psychologic variables as predictors of pain and functional impairment in osteoarthritis of the knee or hip.
      • Cicuttini F.M.
      • Baker J.
      • Hart D.J.
      • Spector T.D.
      Association of pain with radiological changes in different compartments and views of the knee joint.
      • Wluka A.E.
      • Wolfe R.
      • Stuckey S.
      • Cicuttini F.M.
      How does tibial cartilage volume relate to symptoms in subjects with knee osteoarthritis?.
      . One of the reasons for this apparent discrepancy may be the definition of knee OA. Kellgren Lawrence (KL) grading is the conventional system most used to grade the radiographic severity of knee OA
      , but in this categorical system, joint space narrowing is not assessed separately. A recent cross-sectional study showed that the association between joint space narrowing and osteophytosis was not as high as expected on plain radiographs
      • Muraki S.
      • Oka H.
      • Akune T.
      • En-yo Y.
      • Yoshida M.
      • Suzuki T.
      • et al.
      Independent association of joint space narrowing and osteophyte formation at the knee with health-related quality of life in Japan: a cross-sectional study.
      . In addition, joint space narrowing and osteophytosis had distinct effects on QOL
      • Muraki S.
      • Oka H.
      • Akune T.
      • En-yo Y.
      • Yoshida M.
      • Suzuki T.
      • et al.
      Independent association of joint space narrowing and osteophyte formation at the knee with health-related quality of life in Japan: a cross-sectional study.
      . These accumulating lines of evidence have indicated that joint space narrowing and osteophytosis may have distinct etiologic mechanisms, and their progression may be neither constant nor proportional. Although osteophytosis also has some effect on ADL and QOL
      • Muraki S.
      • Oka H.
      • Akune T.
      • En-yo Y.
      • Yoshida M.
      • Suzuki T.
      • et al.
      Independent association of joint space narrowing and osteophyte formation at the knee with health-related quality of life in Japan: a cross-sectional study.
      , joint space narrowing is the primary outcome in studies of OA
      • Altman R.
      • Brandt K.
      • Hochberg M.
      • Moskowitz R.
      Design and conduct of clinical trials in people with osteoarthritis.
      . Thus, to examine the association between knee OA and pain, joint space narrowing should be assessed separately. Chan et al. examined the association of joint space narrowing and duration of pain in patients with knee OA and found a significant association
      • Chan W.P.
      • Huang G.S.
      • Hsu S.M.
      • Chang Y.C.
      • Ho W.P.
      Radiographic joint space narrowing in osteoarthritis of the knee: relationship to meniscal tears and duration of pain.
      , but joint space narrowing was defined by categorical methods. Because categorical methods are statistically less powerful than continuous methods, the association between pain and knee OA might have been underestimated. To overcome this, a fully automatic system that can quantify the joint space width of knee OA on standard radiographs and allows for objective, accurate, and simple assessment of the structural severity of knee OA was developed
      • Oka H.
      • Muraki S.
      • Akune T.
      • Mabuchi A.
      • Suzuki T.
      • Yoshida H.
      • et al.
      Fully automatic quantification of knee osteoarthritis severity on plain radiographs.
      . Thus far, Kinds et al. measured joint space width and found significant associations with clinical outcomes
      • Kinds M.B.
      • Marijnissen A.C.
      • Bijlsma J.W.
      • Boers M.
      • Lafeber F.P.
      • Welsing P.M.
      Quantitative radiographic features of early knee osteoarthritis: development over 5 years and relationship with symptoms in the CHECK cohort.
      , but the threshold of joint space width for clinical outcomes remains unclear.
      Sex differences have been observed in knee OA. The prevalence of knee OA is higher in women than men, and the association of knee pain with knee OA also differs by sex
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      . Thus, the impact of joint space narrowing and osteophytosis on QOL may also differ between the sexes. Obesity is also one of the few established risk factors for knee OA and pain
      • Schouten J.S.A.G.
      A 12 Year Follow Up Study on Osteoarthritis of the Knee in the General Population.
      • Hart D.J.
      • Doyle D.V.
      • Spector T.D.
      Incidence and risk factors for radiographic knee osteoarthritis in middle-aged women: the Chingford Study.
      • Cooper C.
      • Snow S.
      • McAlindon T.E.
      • Kellingray S.
      • Stuart B.
      • Coggon D.
      • et al.
      Risk factors for the incidence and progression of radiographic knee osteoarthritis.
      • Altman R.D.
      • Fries J.F.
      • Bloch D.A.
      • Carstens J.
      • Cooke T.D.
      • Genant H.
      • et al.
      Radiographic assessment of progression in osteoarthritis.
      • Coggon D.
      • Reading I.
      • Croft P.
      • McLaren M.
      • Barrett D.
      • Cooper C.
      Knee osteoarthritis and obesity.
      • D'Arcy Y.
      Pain and obesity.
      • MacFarlane G.
      • de Silva V.
      • Jones G.
      The relationship between body mass index across the life course and knee pain in adulthood: results from the 1958 birth cohort study.
      . This suggests that a distinct association of joint space narrowing with pain may be found in subjects with and without obesity. However, to the best of our knowledge, there are no population-based studies that assess the effect of obesity on the association of joint space narrowing with pain.
      Therefore, the objective of this study was to clarify the association of joint space narrowing with pain at the knee among Japanese men and women using a fully automatic system to measure joint space width in a large-scale, population-based cohort from the Research on Osteoarthritis/osteoporosis Against Disability (ROAD) study. Furthermore, the threshold of minimum joint space width (mJSW) or medial mJSW/lateral mJSW for pain was determined using receiver operating characteristic (ROC) curve analysis.

      Subjects and methods

      Subjects

      The ROAD study is a nationwide prospective study designed to establish epidemiologic indices for the evaluation of clinical evidence for the development of a disease-modifying treatment for bone and joint diseases (with OA and osteoporosis as the representative bone and joint diseases). It consists of population-based cohorts in several communities in Japan. A detailed profile of the ROAD study has been reported elsewhere
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      • Yoshimura N.
      • Muraki S.
      • Oka H.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of knee osteoarthritis, lumbar spondylosis and osteoporosis in Japanese men and women: the research on Osteoarthritis/Osteoporosis Against Disability (ROAD).
      • Yoshimura N.
      • Muraki S.
      • Oka H.
      • Kawaguchi H.
      • Nakamura K.
      • Akune T.
      Cohort profile: research on osteoarthritis/osteoporosis against disability study.
      , and, thus, only a brief summary is provided here. To date, we have completed the creation of a baseline database including clinical and genetic information for 3040 inhabitants (1061 men and 1,979 women) ranging in age from 23 to 95 years (mean, 70.3 years), who were recruited from resident registration listings in three communities: an urban region in Itabashi, Tokyo, a mountainous region in Hidakagawa, Wakayama, and a coastal region in Taiji, Wakayama. All participants provided their written, informed consent, and the study was conducted with the approval of the ethics committees of the University of Tokyo and the Tokyo Metropolitan Institute of Gerontology. Anthropometric measurements, including height and weight, were taken, and body mass index (BMI; weight [kg]/height2 [m2]) was calculated. Furthermore, all participants were also interviewed by well-experienced orthopedists regarding pain in both knees, by asking: “Have you experienced right knee pain on most days in the past month, in addition to now?” and “Have you experienced left knee pain on most days in the past month, in addition to now?”. Subjects who answered “yes” were defined as having knee pain. Among the 3040 subjects who participated in the baseline study, 30 (1.0%) who underwent unilateral or total knee arthroplasty were excluded. In addition, 35 (1.1%) whose radiographic examinations were insufficient for measuring joint space width, and 195 (6.4%) with lateral knee OA were excluded. One reason for excluding lateral knee OA is that most knee OA in Japan is medial type
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      . The other reason is that medial and lateral knee OA have distinct characteristics, and joint space narrowing occurs in the medial compartment in medial knee OA, but medial joint space width may not change or be larger in lateral knee OA. Furthermore, 47 patients (1.5%) who provided incomplete questionnaires regarding pain and so on were excluded, leaving a total of 2733 (89.9%) subjects (975 men and 1758 women).

      Radiographic assessment

      All participants underwent radiographic examinations of both knees using an anterior-posterior view with weight-bearing and foot map positioning by experienced radiological technologists. The beam was positioned parallel to the floor with no angle and aimed at the joint space. To visualize the joint space properly and to centralize the patella over the lower end of the femur, fluoroscopic guidance with an anterior-posterior X-ray beam was used, and the images were downloaded into Digital Imaging and Communication in Medicine (DICOM) format files. Knee radiographs were read without knowledge of participant clinical status by a single experienced orthopedist (S.M.) using the KL radiographic atlas for overall knee radiographic grades
      , and knee OA was defined as KL grade 2 or greater. The KOACAD system was used to measure mJSW in the medial compartment and OPA at the medial tibia
      • Oka H.
      • Muraki S.
      • Akune T.
      • Mabuchi A.
      • Suzuki T.
      • Yoshida H.
      • et al.
      Fully automatic quantification of knee osteoarthritis severity on plain radiographs.
      , and the knee with the lower mJSW was defined as the designated knee for a participant. The KOACAD system is a fully automatic system that can quantify the major features of knee OA on standard radiographs and allows for objective, accurate, and simple assessment of the structural severity of knee OA in general clinical practice. This system was programmed to measure mJSW in the medial and lateral compartments using digitized knee radiographs. The KOACAD system has been described in detail elsewhere
      • Oka H.
      • Muraki S.
      • Akune T.
      • Mabuchi A.
      • Suzuki T.
      • Yoshida H.
      • et al.
      Fully automatic quantification of knee osteoarthritis severity on plain radiographs.
      • Oka H.
      • Muraki S.
      • Akune T.
      • Nakamura K.
      • Kawaguchi H.
      • Yoshimura N.
      Normal and threshold values of joint space width, joint space area, osteophyte area and fibro-tibial angle using a computer-assisted measuring system (KOACAD) to evaluate knee osteoarthritis: the ROAD study.
      . The KOACAD system was applied to the DICOM data by the experienced orthopedist who developed this system (H.O.); measurement reliability has been shown to be good
      • Oka H.
      • Muraki S.
      • Akune T.
      • Mabuchi A.
      • Suzuki T.
      • Yoshida H.
      • et al.
      Fully automatic quantification of knee osteoarthritis severity on plain radiographs.
      , and the intraclass coefficient of correlation for medial mJSW measured on radiographs for an individual with weight-bearing and foot map positioning was 0.96. Reference values for OPA and mJSW by sex and age strata in Japan using the KOACAD system have been published previously
      • Oka H.
      • Muraki S.
      • Akune T.
      • Nakamura K.
      • Kawaguchi H.
      • Yoshimura N.
      Normal and threshold values of joint space width, joint space area, osteophyte area and fibro-tibial angle using a computer-assisted measuring system (KOACAD) to evaluate knee osteoarthritis: the ROAD study.
      . Lateral knee OA was defined as KL grade 2 or greater with lower lateral mJSW than medial mJSW.

      Statistical analysis

      Differences in age, height, weight, BMI, mJSW, and medial/lateral mJSW between men and women and between subjects with and without pain were examined by the non-paired Student's t-test. The prevalence of knee OA was compared between men and women by the χ2 test. Associations of age, BMI, mJSW, and medial/lateral mJSW with knee pain were determined using multiple logistic regression analysis after adjustment for age, sex, and BMI overall, and after adjustment for age and BMI in men and women. In addition, subjects were classified according to mJSW (<1 mm, ≥1–<2 mm, ≥2–<3 mm, ≥3–<4 mm, ≥4 mm), and the associations of mJSW<1 mm, ≥1–<2 mm, ≥2–<3 mm, and ≥3–<4 mm with pain were determined using multiple logistic regression analysis after adjustment for age and BMI, compared with mJSW≥4 mm. To clarify the effect of BMI on the association of mJSW with pain, subjects were further classified into 10 groups according to mJSW and BMI (BMI <23 kg/m2: mJSW<1 mm, ≥1–<2 mm, ≥2–<3 mm, ≥3–<4 mm, ≥4 mm; BMI ≥ 23 kg/m2: mJSW<1 mm, ≥1–<2 mm, ≥2–<3 mm, ≥3–<4 mm, ≥4 mm), and the association with pain was determined using multiple logistic regression analysis after adjustment for age, compared with BMI <23 kg/m2 and mJSW ≥4 mm. Subjects were also classified according to medial/lateral mJSW (<30%, ≥30–<40%, ≥40–<50%, ≥50–<60%, ≥60–<70%, ≥70–<80%, ≥80%), and the associations of medial/lateral mJSW<30%, ≥30–<40%, ≥40–<50%, ≥50–<60%, ≥60–<70%, and ≥70–<80% with pain were determined using multiple logistic regression analysis after adjustment for age and BMI, compared with medial/lateral mJSW ≥80%. The thresholds of mJSW and medial/lateral mJSW for pain were determined using ROC curve analysis. Data analyses were performed using SAS version 9.0 (SAS Institute Inc., Cary, NC).

      Results

      The characteristics of the 2733 participants in the present study are shown in Table I. The prevalence of knee OA was significantly higher in women than in men. The mJSW and medial mJSW/lateral mJSW were significantly lower in women than in men. The participants in the present study were significantly younger than the non-participants (P < 0.05), while BMI was not significantly different between them (non-participants: age, 74.3 ± 7.9 years; BMI, 23.1 ± 3.2 kg/m2).
      Table ICharacteristics of the subjects in the present study
      OverallMenWomenP values
      N27339751758
      Age, years69.9 ± 11.270.8 ± 10.869.4 ± 11.40.0012
      Height, cm154.4 ± 8.9162.5 ± 6.7150.0 ± 6.4<0.0001
      Weight, kg55.1 ± 10.361.4 ± 10.051.6 ± 8.7<0.0001
      BMI, kg/m223.0 ± 3.323.2 ± 3.022.9 ± 3.50.0493
      Right knee
       Knee OA, %45.333.651.8<0.0001
       mJSW, mm2.8 ± 1.03.2 ± 0.92.6 ± 0.9<0.0001
       medial mJSW/lateral mJSW, %68.7 ± 30.171.1 ± 22.267.4 ± 33.60.0007
      Left knee
       Knee OA, %47.535.854<0.0001
       mJSW, mm2.9 ± 1.03.3 ± 01.02.7 ± 0.9<0.0001
       medial mJSW/lateral mJSW, %70.8 ± 26.373.9 ± 22.769.1 ± 28.0<0.0001
      Except where indicated otherwise, values are means ± SD.
      Knee OA was defined as Kellgren Lawrence grade 2 or worse.
      BMI, body mass index; OA, osteoarthritis; mJSW, minimum joint space width.
      Table II shows age, BMI, mJSW, and medial/lateral mJSW in subjects with and without pain. For the right knee, overall and in women, subjects with pain were older and had higher BMI, narrower mJSW, and smaller medial/lateral mJSW than those without pain. In men, subjects with pain had higher BMI, narrower mJSW, and smaller medial/lateral mJSW than subjects without pain, while age was not significantly different in men with and without pain. For the left knee, results were similar except for age in men. Associations of mJSW and medial/lateral mJSW with right and left knee pain were next examined using multiple logistic regression analysis after adjustment for age, sex, and BMI overall, and after adjustment for age and BMI in men and women (Table II). Odds ratios (ORs) of mJSW (1-mm decrease) for pain were higher than 2, and the ORs of medial/lateral mJSW (10% decrease) for pain were 1.2–1.3.
      Table IIAssociations of age, BMI, mJSW, and medial mJSW/lateral mJSW with knee pain
      Right kneeLeft knee
      Pain +Pain −Adjusted OR95% CIP valuesPain +Pain −Adjusted OR95% CIP values
      Overall
       N
       Age, years72.4 ± 8.669.3 ± 11.6*1.011.00–1.030.049972.8 ± 8.469.3 ± 11.6*1.010.99–1.020.3226
       BMI, kg/m224.4 ± 3.622.7 ± 3.2*1.121.08–1.16<0.000124.2 ± 3.522.8 ± 3.2*1.101.06–1.14<0.0001
       mJSW, mm (1-mm decrease)2.1 ± 1.12.9 ± 0.9*2.171.92–2.50<0.00012.2 ± 1.12.9 ± 0.9*2.221.96–2.56<0.0001
       medial mJSW/lateral mJSW, % (10% decrease)54.3 ± 30.771.9 ± 29.0*1.301.24–1.37<0.000156.6 ± 30.471.0 ± 29.5*1.221.16–1.29<0.0001
      Men
       N
       Age, years71.7 ± 9.470.7 ± 11.00.990.96–1.020.509572.8 ± 8.770.5 ± 11.0*0.980.95–1.010.2278
       BMI, kg/m224.2 ± 3.023.0 ± 3.0*1.091.01–1.180.020724.1 ± 3.123.0 ± 3.0*1.080.995–1.170.0635
       mJSW, mm (1-mm decrease)2.4 ± 1.23.3 ± 0.9*2.331.82–3.03<0.00012.6 ± 1.13.2 ± 0.9*2.501.92–3.23<0.0001
       medial mJSW/lateral mJSW, % (10% decrease)57.2 ± 27.672.8 ± 20.8*1.331.19–1.49<0.000161.3 ± 29.672.3 ± 20.8*1.281.15–1.43<0.0001
      Women
       N
       Age, years72.6 ± 8.468.5 ± 12.0*1.021.005–1.040.013872.7 ± 8.368.6 ± 11.9*1.020.997–1.040.0964
       BMI, kg/m224.4 ± 3.722.5 ± 3.3*1.121.08–1.17<0.000124.3 ± 3.622.6 ± 3.4*1.111.06–1.16<0.0001
       mJSW, mm (1-mm decrease)2.0 ± 1.12.8 ± 0.8*2.131.82–2.50<0.00012.1 ± 1.12.7 ± 0.8*2.171.85–2.56<0.0001
      medial mJSW/lateral mJSW, % (10% decrease)52.2 ± 31.571.4 ± 33.0*1.301.22–1.38<0.000155.1 ± 30.670.3 ± 33.7*1.211.14–1.28<0.0001
      P < 0.05 by non-paired Student's t-test.
      Adjusted ORs were calculated by multiple logistic regression analysis after adjustment for age, sex, and BMI overall and after adjustment for age and BMI in men and women.
      BMI, body mass index; mJSW, minimum joint space width; OR, odds ratio; CI, confidence interval.
      Subjects were then classified according to mJSW (<1 mm, ≥1–<2 mm, ≥2–<3 mm, ≥3–<4 mm, ≥4 mm), and the prevalence of knee pain was examined (Fig. 1, Supplementary Table I). The prevalence of knee pain was more than 60% in subjects with mJSW<1 mm, while it was less than 10% in those with mJSW ≥4 mm. The OR for pain was also calculated after adjustment for age and BMI. Men with mJSW <1 mm, ≥1–<2 mm, and ≥2–<3 mm had significantly higher rates of pain than those with mJSW ≥4 mm, but men with mJSW ≥3–<4 mm did not (Table III). The OR for pain in men with mJSW <1 mm was around 40. Women with mJSW <1 mm and ≥1–<2 mm had significantly higher rates of pain than those with mJSW ≥4 mm, but, women with mJSW ≥2–<3 mm and ≥3–<4 mm did not. The ORs for pain in women with mJSW <1 mm were 12–14. Subjects were further classified into 10 groups according to BMI and mJSW (BMI < 23 kg/m2: mJSW < 1 mm, ≥1–<2 mm, ≥2–<3 mm, ≥3–<4 mm, ≥4 mm; BMI ≥ 23 kg/m2: mJSW < 1 mm, ≥1–<2 mm, ≥2–<3 mm, ≥3–<4 mm, ≥4 mm), and the ORs for pain were calculated (Supplementary Table II). In men, mJSW<1 mm and ≥1–<2 mm with BMI <23 kg/m2 and mJSW < 1 mm, ≥1–<2 mm, and ≥2–<3 mm with BMI ≥23 kg/m2 were significantly associated with pain compared with mJSW ≥4 mm with BMI <23 kg/m2. In women at the right knee, mJSW < 1 mm with BMI <23 kg/m2 and mJSW ≥0–<1 mm and ≥1–<2 mm with BMI ≥23 kg/m2 were significantly associated with pain compared with mJSW ≥4 mm with BMI <23 kg/m2. In women at the left knee, mJSW < 1 mm and ≥1–<2 mm with BMI <23 kg/m2 and mJSW < 1 mm, ≥1–<2 mm, and ≥2–<3 mm with BMI ≥23 kg/m2 were significantly associated with pain compared with mJSW ≥4 mm with BMI <23 kg/m2.
      Figure thumbnail gr1
      Fig. 1Prevalence of knee pain by mJSW (mm) in men and women. mJSW, minimum joint space width.
      Table IIIOR of knee pain by medial mJSW
      MenWomen
      Crude OR95% CIAdjusted OR95% CICrude OR95% CIAdjusted OR95% CI
      Right knee
       <1 mm39.011.9–146.439.411.6–151.818.47.7–51.912.35.0–35.3
       ≥1–<2 mm9.03.8–23.98.53.5–23.07.43.3–19.75.92.6–15.9
       ≥2–<3 mm2.91.3–7.231.4–7.620.9–5.21.80.8–4.9
       ≥3–<4 mm1.30.6–3.30.80.2–3.21.20.5–3.31.30.6–3.5
       ≥4 mm1111
      Left knee
       <1 mm45.514.9–163.338.111.9–142.122.79.4–64.314.05.7–40.2
       ≥1–<2 mm8.74.0–20.28.53.8–20.17.33.3–19.55.32.3–14.2
       ≥2–<3 mm2.31.2–5.02.31.2–5.12.41.1–6.42.00.9–5.3
       ≥3–<4 mm0.90.4–1.90.90.4–1.91.40.6–3.81.40.6–3.7
       ≥4 mm1111
      Adjusted ORs were calculated by multiple logistic regression analysis after adjustment for age and BMI.
      mJSW, minimum joint space width; CI, confidence interval.
      Subjects were also classified according to medial/lateral mJSW (<30%, ≥30–<40%, ≥40–<50%, ≥50–<60%, ≥60–<70%, ≥70–<80%, ≥80%), and the prevalence of knee pain was examined (Fig. 2, Supplementary Table III). The prevalence of knee pain was approximately 60% in subjects with medial mJSW/lateral mJSW < 30%, while it was approximately 10% in those with medial mJSW/lateral mJSW ≥80%. The ORs for pain were also calculated after adjustment for age and BMI (Table IV). Men and women with mJSW <30%, ≥30–<40%, ≥40–<50%, and ≥50–<60% had higher rates of pain compared with those with mJSW ≥80%, except for men with mJSW≥50–<60% at the right knee. The ORs for pain in men with mJSW <30% were 14–20. The OR for pain in women with mJSW <30% was around 10.
      Figure thumbnail gr2
      Fig. 2Prevalence of knee pain by medial mJSW/lateral mJSW (%) in men and women. mJSW, minimum joint space width.
      Table IVOR of knee pain by the ratio of medial mJSW to lateral mJSW
      MenWomen
      Crude OR95% CIAdjusted OR95% CICrude OR95% CIAdjusted OR95% CI
      Right knee
      <30%16.06.9–38.214.56.1–35.014.89.4–23.79.86.1–16.0
      ≥30–<40%5.01.9–12.24.41.6–10.95.73.2–10.04.22.3–7.6
      ≥40–<50%3.11.5–6.52.71.2–5.73.11.9–4.92.41.5–4.0
      ≥50–<60%1.50.7–3.01.40.7–2.92.21.5–3.32.11.4–3.2
      ≥60–<70%1.50.8–2.81.50.8–2.81.40.9–2.01.30.9–2.0
      ≥70–<80%0.90.5–1.90.90.5–1.91.20.8–1.81.10.7–1.8
      ≥80%1111
      Left knee
      <30%23.010.7–51.318.98.6–43.115.59.8–25.010.36.4–16.9
      ≥30–<40%6.12.2–15.85.72.0–14.83.62.0–6.42.81.5–5.0
      ≥40–<50%4.01.7–8.83.61.5–8.05.43.4–8.64.32.7–7.0
      ≥50–<60%2.21.1–4.42.11.0–4.12.61.7–3.82.31.5–3.4
      ≥60–<70%1.20.6–2.41.20.6–2.41.50.97–2.21.40.9–2.1
      ≥70–<80%1.00.5–1.91.00.5–2.00.80.5–1.30.80.5–1.3
      ≥80%1111
      Adjusted ORs were calculated by multiple logistic regression analysis after adjustment for age and BMI.
      mJSW, minimum joint space width; CI, confidence interval.
      The threshold values of mJSW for knee pain were then determined using ROC curve analysis (Supplementary Fig. 1). In men, the threshold values of mJSW for pain at the right and left knees were 2.87 mm (sensitivity 0.67, specificity 0.65, AUC 0.70, 95% confidence interval (CI) 0.64–0.75) and 2.82 mm (sensitivity 0.62, specificity 0.67, AUC 0.72, 95% CI 0.66–0.77), respectively. In women, the threshold values of mJSW for pain at the right and left knees were 2.01 mm (sensitivity 0.43, specificity 0.689, AUC 0.69, 95% CI 0.66–0.73) and 2.44 mm (sensitivity 0.59, specificity 0.75, AUC 0.71, 95% CI 0.67–0.74), respectively. Threshold values of medial/lateral mJSW for knee pain were also determined using ROC curve analysis (Supplementary Fig. 2). In men, the threshold values of medial/lateral mJSW for pain at the right and left knees were 55.2% (sensitivity 0.45, specificity 0.68, AUC 0.66, 95% CI 0.60–0.72) and 57.9% (sensitivity 0.49, specificity 0.84, AUC 0.70, 95% CI 0.64–0.76), respectively. In women, the threshold values of medial/lateral mJSW for pain at the right and left knees were 57.9% (sensitivity 0.57, specificity 0.75, AUC 0.69, 95% CI 0.66–0.73) and 57.7% (sensitivity 0.58, specificity 0.76, AUC 0.71, 95% CI 0.68–0.74), respectively.

      Discussion

      Joint space narrowing is the primary outcome in studies of knee OA
      • Altman R.D.
      • Fries J.F.
      • Bloch D.A.
      • Carstens J.
      • Cooke T.D.
      • Genant H.
      • et al.
      Radiographic assessment of progression in osteoarthritis.
      , because cartilage damage, which is one of the main causes of knee symptoms, is seen as a smaller mJSW
      • Kinds M.B.
      • Marijnissen A.C.
      • Bijlsma J.W.
      • Boers M.
      • Lafeber F.P.
      • Welsing P.M.
      Quantitative radiographic features of early knee osteoarthritis: development over 5 years and relationship with symptoms in the CHECK cohort.
      . Previous studies have shown significant associations of joint space narrowing with pain
      • Chan W.P.
      • Huang G.S.
      • Hsu S.M.
      • Chang Y.C.
      • Ho W.P.
      Radiographic joint space narrowing in osteoarthritis of the knee: relationship to meniscal tears and duration of pain.
      • Kinds M.B.
      • Marijnissen A.C.
      • Bijlsma J.W.
      • Boers M.
      • Lafeber F.P.
      • Welsing P.M.
      Quantitative radiographic features of early knee osteoarthritis: development over 5 years and relationship with symptoms in the CHECK cohort.
      , though the threshold of joint space width for pain remained unclear. This is the first study to clarify the effect of joint space narrowing on knee pain using a large-scale, population-based, cohort study. In addition, joint space narrowing was evaluated not by categorical grade but by continuous values, using mJSW at the knee. In the present study, mJSW < 3 mm in men and mJSW < 2 mm in women were significantly associated with knee pain, compared with mJSW ≥4 mm, and the OR of mJSW < 1 mm for knee pain was quite high, particularly in men. It was also found that the effect of mJSW on pain was affected by BMI. Medial mJSW/lateral/mJSW < 60% was also significantly associated with knee pain in men and women, compared with medial mJSW/lateral mJSW ≥ 80%. Using ROC curve analysis, the thresholds of mJSW in men and women were found to be approximately 3 mm in men and 2 mm in women, while those of medial mJSW/lateral mJSW were approximately 60% in both men and women.
      Although much effort has been devoted toward a definition of knee pain, the correlation with radiographic severity of knee OA was not as strong as one would expect
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      • Summers M.N.
      • Haley W.E.
      • Reveille J.D.
      • Alarcon G.S.
      Radiographic assessment and psychologic variables as predictors of pain and functional impairment in osteoarthritis of the knee or hip.
      • Cicuttini F.M.
      • Baker J.
      • Hart D.J.
      • Spector T.D.
      Association of pain with radiological changes in different compartments and views of the knee joint.
      • Wluka A.E.
      • Wolfe R.
      • Stuckey S.
      • Cicuttini F.M.
      How does tibial cartilage volume relate to symptoms in subjects with knee osteoarthritis?.
      . In fact, our previous study showed that the OR of severe knee OA defined as KL grade 3 or 4 for knee pain was 8.6 in men and 4.4 in women
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      , which was significant, but the OR was not as high as expected. One of the reasons for this is that knee pain may arise from a variety of structures other than joint cartilage, such as menisci, synovium, ligaments, bursae, bone, and bone marrow
      • Saito T.
      • Koshino T.
      Distribution of neuropeptides in synovium of the knee with osteoarthritis.
      • Bollet A.J.
      Edema of the bone marrow can cause pain in osteoarthritis and other diseases of bone and joints.
      • Teichtahl A.J.
      • Wluka A.E.
      • Morris M.E.
      • Davis S.R.
      • Cicuttini F.M.
      The relationship between the knee adduction moment and knee pain in middle-aged women without radiographic osteoarthritis.
      • Thorp L.E.
      • Sumner D.R.
      • Wimmer M.A.
      • Block J.A.
      Relationship between pain and medial knee joint loading in mild radiographic knee osteoarthritis.
      • Felson D.T.
      • Niu J.
      • Guermazi A.
      • Roemer F.
      • Aliabadi P.
      • Clancy M.
      • et al.
      Correlation of the development of knee pain with enlarging bone marrow lesions on magnetic resonance imaging.
      . Another reason may be due to the definition of knee OA. Knee OA is characterized by the pathological features of joint space narrowing and osteophytosis. However, most conventional systems for grading radiographic severity have been categorical grades, such as KL grading
      , which cannot assess joint space narrowing individually. Several studies have shown that knee OA has a strong effect on QOL
      • Woo J.
      • Lau E.
      • Lee P.
      • Kwok T.
      • Lau W.C.
      • Chan C.
      • et al.
      Impact of osteoarthritis on quality of life in a Hong Kong Chinese population.
      • Brazier J.E.
      • Harper R.
      • Munro J.
      • Walters S.J.
      • Snaith M.L.
      Generic and condition-specific outcome measures for people with osteoarthritis of the knee.
      • Hill C.L.
      • Parsons J.
      • Taylor A.
      • Leach G.
      Health related quality of life in a population sample with arthritis.
      • Muraki S.
      • Akune T.
      • Oka H.
      • En-yo Y.
      • Yoshida M.
      • Saika A.
      • et al.
      Association of radiographic and symptomatic knee osteoarthritis with health-related quality of life in a population-based cohort study in Japan: the ROAD study.
      , but in these studies, knee OA was defined by categorical grades such as KL grade or American College of Rheumatology (ACR) grade, total knee arthroplasty, and self-questionnaire. In addition, joint space narrowing was separately evaluated using a radiographic atlas of individual features published by the Osteoarthritis Research Society International (OARSI) in 1995
      • Altman R.D.
      • Hochberg M.
      • Murphy Jr., W.A.
      • Wolfe F.
      • Lequesne M.
      Atlas of individual radiographic features in osteoarthritis.
      and revised in 2007
      • Altman R.D.
      • Gold G.E.
      Atlas of individual radiographic features in osteoarthritis, revised.
      . Chan et al. examined the association of joint space narrowing and duration of pain in patients with knee OA using categorical methods
      • Chan W.P.
      • Huang G.S.
      • Hsu S.M.
      • Chang Y.C.
      • Ho W.P.
      Radiographic joint space narrowing in osteoarthritis of the knee: relationship to meniscal tears and duration of pain.
      . However, the grading is still limited in reproducibility and sensitivity due to the subjective judgment of individual observers and the categorical classification. Furthermore, because categorical methods are statistically less powerful than continuous methods, the association between pain and knee OA might have been underestimated in previous studies. Kinds et al. measured joint space width and found significant associations with clinical outcomes
      • Kinds M.B.
      • Marijnissen A.C.
      • Bijlsma J.W.
      • Boers M.
      • Lafeber F.P.
      • Welsing P.M.
      Quantitative radiographic features of early knee osteoarthritis: development over 5 years and relationship with symptoms in the CHECK cohort.
      , but the threshold of joint space width for clinical outcomes remained unclear. In the present study, to overcome this problem, joint space width was evaluated using a fully automatic system, and the OR of mJSW <1 mm for knee pain was quite high, particularly in men, and it was possible to establish the threshold values of mJSW for knee pain, which may indicate that mJSW is better for diagnosing knee OA than KL grade. In the present study, 6% of men with mJSW <3 mm and 14% of women with mJSW <2 mm, which were the threshold values in the present study, had knee pain. In addition, our previous study showed that 10% of men without knee OA and 20% of women without knee OA had knee pain
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      . These subjects have knee pain, despite having no radiographical changes. This indicates that at least 10% and 20% of knee pain in men and women, respectively, may be explained by factors other than radiographical changes.
      In the present study, sex differences were found in the association of mJSW with pain. These discrepancies between the sexes are explained by several factors. First, women are more susceptible to pain than men
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      . In fact, our previous study showed that the OR for knee pain in women without radiographic knee OA (KL = 0/1) was greater than that in men without radiographic knee OA
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      . In the present study, the prevalence of knee pain was 5–6% in men with mJSW ≥4 mm, while it was 7–9% in women with mJSW ≥4 mm. This high prevalence of knee pain in women with mJSW ≥4 mm, which are reference data, may partly explain the lower OR for knee pain in women than men. Second, men with normal knees had wider joint space widths than women with normal knees. Our previous study showed that mean mJSW in men with KL = 0 was approximately 4 mm, while that in women with KL = 0 was approximately 3 mm
      • Oka H.
      • Muraki S.
      • Akune T.
      • Nakamura K.
      • Kawaguchi H.
      • Yoshimura N.
      Normal and threshold values of joint space width, joint space area, osteophyte area and fibro-tibial angle using a computer-assisted measuring system (KOACAD) to evaluate knee osteoarthritis: the ROAD study.
      . This means that subjects with mJSW = 3 mm have a normal knee in women, while they have joint space narrowing at the knee in men. In addition, mJSW = 1 mm means 75% cartilage loss in men, while it represent 67% cartilage loss in women. In fact, the associations of medial mJSW/lateral mJSW with pain were similar in men and women, which may also explain the lower OR for knee pain in women than men.
      Obesity is one of the few established risk factors for knee OA and pain
      • Schouten J.S.A.G.
      A 12 Year Follow Up Study on Osteoarthritis of the Knee in the General Population.
      • Hart D.J.
      • Doyle D.V.
      • Spector T.D.
      Incidence and risk factors for radiographic knee osteoarthritis in middle-aged women: the Chingford Study.
      • Cooper C.
      • Snow S.
      • McAlindon T.E.
      • Kellingray S.
      • Stuart B.
      • Coggon D.
      • et al.
      Risk factors for the incidence and progression of radiographic knee osteoarthritis.
      • Altman R.D.
      • Fries J.F.
      • Bloch D.A.
      • Carstens J.
      • Cooke T.D.
      • Genant H.
      • et al.
      Radiographic assessment of progression in osteoarthritis.
      • Coggon D.
      • Reading I.
      • Croft P.
      • McLaren M.
      • Barrett D.
      • Cooper C.
      Knee osteoarthritis and obesity.
      • D'Arcy Y.
      Pain and obesity.
      • MacFarlane G.
      • de Silva V.
      • Jones G.
      The relationship between body mass index across the life course and knee pain in adulthood: results from the 1958 birth cohort study.
      . A clinical review article reported that 69% of knee replacements in middle-aged females can be attributed to obesity
      • D'Arcy Y.
      Pain and obesity.
      , and it has been estimated that, if overweight and obese individuals reduced their weight to reach normal BMIs, about 50% of knee OA cases would be eliminated
      • Coggon D.
      • Reading I.
      • Croft P.
      • McLaren M.
      • Barrett D.
      • Cooper C.
      Knee osteoarthritis and obesity.
      . However, to the best of our knowledge, there are no population-based studies that assess the effect of obesity on the association of joint space narrowing with pain. In the present study, a distinct effect of BMI was found on the association of mJSW with pain. For example, at the right knee in women, mJSW ≥1–<2 mm in women with BMI ≥23 kg/m2 was significantly associated with pain, while mJSW ≥1–<2 mm in women with BMI <23 kg/m2 was not, compared to mJSW≥4 mm with BMI <23 kg/m2. In addition, the OR was similar between mJSW ≥1–<2 mm with BMI ≥23 kg/m2 and mJSW <1 mm with BMI <23 kg/m2. These indicate that weight loss may be an effective way to reduce knee pain even when joint space narrowing is present at the knee.
      There are limitations in the present study. First, this was a large-scale, population-based, cross-sectional study of baseline data. Thus, causal relationships could not be determined. The ROAD study is a longitudinal survey, so further progress may help elucidate any causal relationships. Second, the threshold in the present study was calculated by a particular statistical method, but certain situations may favor sensitivity over specificity, e.g., screening. In addition, the sensitivity and specificity were modest in the present study. These may be partly explained by the fact that knee pain can arise from a variety of structures other than joint cartilage, such as menisci, synovium, ligaments, bursae, bone, and bone marrow
      • Saito T.
      • Koshino T.
      Distribution of neuropeptides in synovium of the knee with osteoarthritis.
      • Bollet A.J.
      Edema of the bone marrow can cause pain in osteoarthritis and other diseases of bone and joints.
      • Teichtahl A.J.
      • Wluka A.E.
      • Morris M.E.
      • Davis S.R.
      • Cicuttini F.M.
      The relationship between the knee adduction moment and knee pain in middle-aged women without radiographic osteoarthritis.
      • Thorp L.E.
      • Sumner D.R.
      • Wimmer M.A.
      • Block J.A.
      Relationship between pain and medial knee joint loading in mild radiographic knee osteoarthritis.
      • Felson D.T.
      • Niu J.
      • Guermazi A.
      • Roemer F.
      • Aliabadi P.
      • Clancy M.
      • et al.
      Correlation of the development of knee pain with enlarging bone marrow lesions on magnetic resonance imaging.
      , which are unable to be assessed radiologically. However, using the KOACAD system, it was possible to demonstrate strong associations of mJSW with knee pain and to establish the threshold of mJSW for knee pain, which may indicate that mJSW is more useful than categorical methods for diagnosing knee OA. Third, cases with lateral knee OA were excluded, leading to a selective sample. One reason for excluding lateral knee OA is that most knee OA in Japan is medial type. There are racial differences in the ratio of lateral to medial knee OA, and previous studies showed that the ratio of lateral to medial knee OA was 0.20 in Caucasian and 0.64 in Chinese populations
      • Felson D.T.
      • Nevitt M.C.
      • Zhang Y.
      • Aliabadi P.
      • Baumer B.
      • Gale D.
      • et al.
      High prevalence of lateral knee osteoarthritis in Beijing Chinese compared with Framingham caucasian asubjects.
      . In the Amsterdam OA Cohort, lateral knee OA is rather common, and it occurs in association with OA features in other knee compartments
      • van der Esch M.
      • Knol D.L.
      • Schaffers I.C.
      • Reiding D.J.
      • van Schaardenburg D.
      • Knoop J.
      • et al.
      Osteoarthritis of the knee: multicompartmental or compartmental disease?.
      . However, our previous study showed that the ratio of lateral to medial knee OA was 0.10 in Japan, which indicates that knee OA was medial type. The other reason for excluding lateral knee OA is that medial and lateral knee OA have distinct characteristics, because, in medial knee OA, there is narrowing of the medial mJSW, while in lateral knee OA, there is narrowing of the lateral mJSW. Thus, the effect of medial mJSW on pain may be obscured by lateral knee OA, because medial joint space width may not change or be larger in lateral knee OA. Thus, the aim of the present study was to clarify the effect of medial knee OA on pain, although excluding lateral OA leads to a selective sample. Lastly, it was not possible to clarify whether the threshold in the present study can apply to other races or populations, because the prevalence of knee OA and the ratio of medial knee OA/lateral knee OA are quite different among races
      • Muraki S.
      • Oka H.
      • Akune T.
      • Mabuchi A.
      • En-yo Y.
      • Yoshida M.
      • et al.
      Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
      • Felson D.T.
      • Nevitt M.C.
      • Zhang Y.
      • Aliabadi P.
      • Baumer B.
      • Gale D.
      • et al.
      High prevalence of lateral knee osteoarthritis in Beijing Chinese compared with Framingham caucasian asubjects.
      • van der Esch M.
      • Knol D.L.
      • Schaffers I.C.
      • Reiding D.J.
      • van Schaardenburg D.
      • Knoop J.
      • et al.
      Osteoarthritis of the knee: multicompartmental or compartmental disease?.
      , and the association of knee OA with pain among them may be quite different. To clarify this, international collaborative studies using the KOACAD system are needed.
      In conclusion, the present cross-sectional study using a large population from the ROAD study showed that joint space narrowing was strongly associated with knee pain. The threshold of mJSW with knee pain was approximately 3 mm in men and 2 mm in women, while the threshold of medial mJSW/lateral mJSW was approximately 60% in both men and women. BMI was found to have a distinct effect on the association of mJSW with pain. Further studies, along with continued longitudinal surveys in the ROAD study, will help improve our understanding of the mechanisms of joint space narrowing at the knee and their relationship with pain.

      Acknowledgments

      This work was supported by Grants-in-Aid for Scientific Research (S19109007, B20390182, C20591737, C20591774), for Young Scientists (A18689031), and for Exploratory Research (19659305) from the Japanese Ministry of Education, Culture, Sports, Science, and Technology, grants (H17-Men-eki-009, H18-Choujyu-037, H20-Choujyu-009, H21-Chouju-Wakate-011 and H22-Chouju-Wakate-007) from the Ministry of Health, Labour and Welfare, a Research Aid from the Japanese Orthopaedic Association (JOA-Subsidized Science Project Research 2006-1), and Grant No.166 from the Japan Orthopaedics and Traumatology Foundation.
      The authors would like to thank Dr Anamizu and members of the Department of Orthopedics; Mr Kutsuma and other members of the Department of Radiology at Tokyo Metropolitan Geriatric Medical Center; Mrs Tomoko Takijiri and other members of the Public Office in Hidakagawa Town; and Mrs Tamako Tsutsumi and Mrs Kanami Maeda, and other members of the Public Office in Taiji Town, for their assistance in locating and scheduling participants for examinations.

      Appendix A. Supplementary data

      The following are the supplementary data related to this article:

      Author contributions

      All authors have made substantial contributions to all three of the following sections:
      • (1)
        the conception and design of the study, or acquisition of data, or analysis and interpretation of data;
      • (2)
        drafting the article or revising it critically for important intellectual content; and
      • (3)
        final approval of the version to be submitted.

      Conflicts of interest

      There are no conflicts of interest.

      References

        • Sharma L.
        • Kapoor D.
        Epidemiology of osteoarthritis.
        in: Moskowitz R.W. Altman R.D. Hochberg M.C. Buckwalter J.A. Goldberg V.M. Osteoarthritis: Diagnosis and Medical/Surgical Management. 4th edn. Lippincott Williams & Wilkins, Philadelphia2007: 3-26
        • Guccione A.A.
        • Felson D.T.
        • Anderson J.J.
        • Anthony J.M.
        • Zhang Y.
        • Wilson P.W.
        • et al.
        The effects of specific medical conditions on the functional limitations of elders in the Framingham Study.
        Am J Public Health. 1994; 84: 351-358
        • Felson D.T.
        • Zhang Y.
        An update on the epidemiology of knee and hip osteoarthritis with a view to prevention.
        Arthritis Rheum. 1998; 41: 1343-1355
        • Muraki S.
        • Oka H.
        • Akune T.
        • Mabuchi A.
        • En-yo Y.
        • Yoshida M.
        • et al.
        Prevalence of radiographic knee osteoarthritis and its association with knee pain in the elderly of Japanese population-based cohorts: the ROAD study.
        Osteoarthritis Cartilage. 2009; 17: 1137-1143
        • Yoshimura N.
        • Muraki S.
        • Oka H.
        • Mabuchi A.
        • En-yo Y.
        • Yoshida M.
        • et al.
        Prevalence of knee osteoarthritis, lumbar spondylosis and osteoporosis in Japanese men and women: the research on Osteoarthritis/Osteoporosis Against Disability (ROAD).
        J Bone Miner Metab. 2009; 27: 620-628
        • Ministry of Health, Labour and Welfare
        The Outline of the Results of National Livelihood Survey.
        2010 (Available at: http://www.mhlw.go.jp/toukei/saikin/hw/k-tyosa/k-tyosa10/index.html)
        • Linaker C.H.
        • Walker-Bone K.
        • Palmer K.
        • Cooper C.
        Frequency and impact of regional musculoskeletal disorders.
        Baillieres Best Pract Res Clin Rheumatol. 1999; 13: 197-215
        • Summers M.N.
        • Haley W.E.
        • Reveille J.D.
        • Alarcon G.S.
        Radiographic assessment and psychologic variables as predictors of pain and functional impairment in osteoarthritis of the knee or hip.
        Arthritis Rheum. 1988; 31: 204-209
        • Cicuttini F.M.
        • Baker J.
        • Hart D.J.
        • Spector T.D.
        Association of pain with radiological changes in different compartments and views of the knee joint.
        Osteoarthritis Cartilage. 1996; 4: 143-147
        • Wluka A.E.
        • Wolfe R.
        • Stuckey S.
        • Cicuttini F.M.
        How does tibial cartilage volume relate to symptoms in subjects with knee osteoarthritis?.
        Ann Rheum Dis. 2004; 63: 264-268
      1. Kellgren J.H. Lawrence J.S. The Epidemiology of Chronic Rheumatism: Atlas of Standard Radiographs of Arthritis. Blackwell Scientific, Oxford1963
        • Muraki S.
        • Oka H.
        • Akune T.
        • En-yo Y.
        • Yoshida M.
        • Suzuki T.
        • et al.
        Independent association of joint space narrowing and osteophyte formation at the knee with health-related quality of life in Japan: a cross-sectional study.
        Arthritis Rheum. 2011; 63: 3859-3864
        • Altman R.
        • Brandt K.
        • Hochberg M.
        • Moskowitz R.
        Design and conduct of clinical trials in people with osteoarthritis.
        Osteoarthritis Cartilage. 1996; 4: 217-243
        • Chan W.P.
        • Huang G.S.
        • Hsu S.M.
        • Chang Y.C.
        • Ho W.P.
        Radiographic joint space narrowing in osteoarthritis of the knee: relationship to meniscal tears and duration of pain.
        Skeletal Radiol. 2008; 37: 917-922
        • Oka H.
        • Muraki S.
        • Akune T.
        • Mabuchi A.
        • Suzuki T.
        • Yoshida H.
        • et al.
        Fully automatic quantification of knee osteoarthritis severity on plain radiographs.
        Osteoarthritis Cartilage. 2008; 16: 1300-1306
        • Kinds M.B.
        • Marijnissen A.C.
        • Bijlsma J.W.
        • Boers M.
        • Lafeber F.P.
        • Welsing P.M.
        Quantitative radiographic features of early knee osteoarthritis: development over 5 years and relationship with symptoms in the CHECK cohort.
        J Rheumatol. 2014; 40: 58-65
        • Schouten J.S.A.G.
        A 12 Year Follow Up Study on Osteoarthritis of the Knee in the General Population.
        ([thesis]) Erasmus University Medical School, Rotterdam1991
        • Hart D.J.
        • Doyle D.V.
        • Spector T.D.
        Incidence and risk factors for radiographic knee osteoarthritis in middle-aged women: the Chingford Study.
        Arthritis Rheum. 1999; 42: 17-24
        • Cooper C.
        • Snow S.
        • McAlindon T.E.
        • Kellingray S.
        • Stuart B.
        • Coggon D.
        • et al.
        Risk factors for the incidence and progression of radiographic knee osteoarthritis.
        Arthritis Rheum. 2000; 43: 995-1000
        • Altman R.D.
        • Fries J.F.
        • Bloch D.A.
        • Carstens J.
        • Cooke T.D.
        • Genant H.
        • et al.
        Radiographic assessment of progression in osteoarthritis.
        Arthritis Rheum. 1987; 30: 1214-1225
        • Coggon D.
        • Reading I.
        • Croft P.
        • McLaren M.
        • Barrett D.
        • Cooper C.
        Knee osteoarthritis and obesity.
        Int J Obes Relat Metab Disord. 2001; 25: 622
        • D'Arcy Y.
        Pain and obesity.
        Nurs Manage. 2012; 43: 21-26
        • MacFarlane G.
        • de Silva V.
        • Jones G.
        The relationship between body mass index across the life course and knee pain in adulthood: results from the 1958 birth cohort study.
        Rheumatology. 2011; 50: 2251-2256
        • Yoshimura N.
        • Muraki S.
        • Oka H.
        • Kawaguchi H.
        • Nakamura K.
        • Akune T.
        Cohort profile: research on osteoarthritis/osteoporosis against disability study.
        Int J Epidemiol. 2010; 39: 988-995
        • Oka H.
        • Muraki S.
        • Akune T.
        • Nakamura K.
        • Kawaguchi H.
        • Yoshimura N.
        Normal and threshold values of joint space width, joint space area, osteophyte area and fibro-tibial angle using a computer-assisted measuring system (KOACAD) to evaluate knee osteoarthritis: the ROAD study.
        J Orthop Sci. 2010; 15: 781-789
        • Saito T.
        • Koshino T.
        Distribution of neuropeptides in synovium of the knee with osteoarthritis.
        Clin Orthop Relat Res. 2000; 376: 172-182
        • Bollet A.J.
        Edema of the bone marrow can cause pain in osteoarthritis and other diseases of bone and joints.
        Ann Intern Med. 2001; 134: 591-593
        • Teichtahl A.J.
        • Wluka A.E.
        • Morris M.E.
        • Davis S.R.
        • Cicuttini F.M.
        The relationship between the knee adduction moment and knee pain in middle-aged women without radiographic osteoarthritis.
        J Rheumatol. 2006; 33: 1845-1848
        • Thorp L.E.
        • Sumner D.R.
        • Wimmer M.A.
        • Block J.A.
        Relationship between pain and medial knee joint loading in mild radiographic knee osteoarthritis.
        Arthritis Rheum. 2007; 57: 1254-1260
        • Felson D.T.
        • Niu J.
        • Guermazi A.
        • Roemer F.
        • Aliabadi P.
        • Clancy M.
        • et al.
        Correlation of the development of knee pain with enlarging bone marrow lesions on magnetic resonance imaging.
        Arthritis Rheum. 2007; 56: 2986-2992
        • Woo J.
        • Lau E.
        • Lee P.
        • Kwok T.
        • Lau W.C.
        • Chan C.
        • et al.
        Impact of osteoarthritis on quality of life in a Hong Kong Chinese population.
        J Rheumatol. 2004; 31: 2433-2438
        • Brazier J.E.
        • Harper R.
        • Munro J.
        • Walters S.J.
        • Snaith M.L.
        Generic and condition-specific outcome measures for people with osteoarthritis of the knee.
        Rheumatology (Oxford). 1999; 38: 870-877
        • Hill C.L.
        • Parsons J.
        • Taylor A.
        • Leach G.
        Health related quality of life in a population sample with arthritis.
        J Rheumatol. 1999; 26: 2029-2035
        • Muraki S.
        • Akune T.
        • Oka H.
        • En-yo Y.
        • Yoshida M.
        • Saika A.
        • et al.
        Association of radiographic and symptomatic knee osteoarthritis with health-related quality of life in a population-based cohort study in Japan: the ROAD study.
        Osteoarthritis Cartilage. 2010; 18: 1227-1234
        • Altman R.D.
        • Hochberg M.
        • Murphy Jr., W.A.
        • Wolfe F.
        • Lequesne M.
        Atlas of individual radiographic features in osteoarthritis.
        Osteoarthritis Cartilage. 1995; 3: 3-70
        • Altman R.D.
        • Gold G.E.
        Atlas of individual radiographic features in osteoarthritis, revised.
        Osteoarthritis Cartilage. 2007; 15: A1-A56
        • Felson D.T.
        • Nevitt M.C.
        • Zhang Y.
        • Aliabadi P.
        • Baumer B.
        • Gale D.
        • et al.
        High prevalence of lateral knee osteoarthritis in Beijing Chinese compared with Framingham caucasian asubjects.
        Arthritis Rheum. 2002; 46: 1217-1222
        • van der Esch M.
        • Knol D.L.
        • Schaffers I.C.
        • Reiding D.J.
        • van Schaardenburg D.
        • Knoop J.
        • et al.
        Osteoarthritis of the knee: multicompartmental or compartmental disease?.
        Rheumatology (Oxford). 2014; 53: 540-546