You are here

Article Text

Article menu
Clinical and epidemiological research
Extended report
The relative contribution of mechanical stress and systemic processes in different types of osteoarthritis: the NEO study
  1. A W Visser1,
  2. R de Mutsert2,
  3. S le Cessie2,3,
  4. M den Heijer2,4,
  5. F R Rosendaal2,5,
  6. M Kloppenburg1,2
  7. for the NEO Study Group
    1. 1Department of Rheumatology, Leiden University Medical Center, Leiden, The Netherlands
    2. 2Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, The Netherlands
    3. 3Department of Medical Statistics and Bio-informatics, Leiden University Medical Center, Leiden, The Netherlands
    4. 4Department of Endocrinology, Vrije Universiteit Medical Center, Amsterdam, The Netherlands
    5. 5Department of Thrombosis and Hemostasis, Leiden University Medical Center, Leiden, The Netherlands
    1. Correspondence to A Willemien Visser, Department of Rheumatology, Leiden University Medical Center, C1-R, P.O. Box 9600, Leiden 2300 RC, The Netherlands; a.w.visser{at}lumc.nl

    Abstract

    Objective To study the relative contribution of surrogates for mechanical stress and systemic processes with osteoarthritis (OA) in weight-bearing and non-weight-bearing joints.

    Methods The Netherlands Epidemiology of Obesity study is a population-based cohort including 6673 participants (range 45–65 years, 56% women, median body mass index 26 kg/m2). Weight (kg) and fat mass (kg) were measured, fat-free mass (kg) was calculated. The metabolic syndrome was defined following the Adult Treatment Panel III criteria. Knee and hand OA were defined according to the American College of Rheumatology clinical criteria.

    Logistic regression analyses were performed to associate surrogates for mechanical stress (such as weight, fat-free mass) and systemic processes (such as metabolic syndrome) with OA in knees alone, knees and hands or hands alone, adjusted for age, sex, height, smoking, education and ethnicity, and when appropriate for metabolic factors and weight.

    Results Knee, knee and hand, and hand OA were present in 10%, 4% and 8% of the participants, respectively. Knee OA was associated with weight and fat-free mass, adjusted for metabolic factors (OR 1.49 (95% CI 1.32 to 1.68) and 2.05 (1.60 to 2.62), respectively). Similar results were found for OA in knees and hands (OR 1.51 (95% CI 1.29 to 1.78) and 2.17 (95% CI 1.52 to 3.10) respectively). Hand OA was associated with the metabolic syndrome, adjusted for weight (OR 1.46 (95% CI 1.06 to 2.02)).

    Conclusions In knee OA, whether or not in co-occurrence with hand OA, surrogates for mechanical stress are suggested to be the most important risk factors, whereas in hand OA alone, surrogates for systemic processes are the most important risk factors.

    • Knee Osteoarthritis
    • Hand Osteoarthritis
    • Epidemiology

    https://doi.org/10.1136/annrheumdis-2013-205012

    Statistics from Altmetric.com

      Request Permissions

      If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

      Introduction

      Overweight and obesity are well-known risk factors for osteoarthritis (OA) in weight-bearing and non-weight-bearing joints.1 ,2 Several mechanisms are thought to explain the association between obesity and OA. First, increased mechanical stress can result in damaged joint tissue.3 Second, systemic processes seem to be involved. Adipose tissue is known as a source of proinflammatory and anti-inflammatory adipokines, which have been suggested to play a role in OA pathogenesis.4–7 Furthermore, hyperglycaemia and diabetes have been related to OA,8–12 possibly via insulin-like growth factor I resistance of chondrocytes,13 via changes in striated muscles due to insulin resistance,14 or via formation of advanced glycation end products.15 ,16 The association of OA with measures of atherosclerosis suggests another systemic link with OA,17 ,18 possibly via systemic inflammation or pathology of subchondral bone vasculature.19

      The relative contribution of mechanical stress and systemic processes to different types of OA remains unclear. In OA of weight-bearing joints as the knees, the concept of excessive mechanical stress leading to OA is supported by previous reported associations of weight or lean body mass with knee OA.20–23 The role of systemic processes in OA of weight-bearing joints is questionable, and difficult to identify since in obese individuals increased mechanical stress and systemic processes frequently occur together. Two recent studies on the metabolic syndrome, as surrogate for systemic processes, in relation to knee OA reported conflicting results. One observed an association between the metabolic syndrome and increased OA incidence even after adjustment for body mass index (BMI), whereas the other did not.24 ,25 So far, no studies examined knee OA in the presence of OA of non-weight-bearing joints as the hands, while this type of polyarticular OA might be particularly driven by systemic processes.

      In hand OA, being non-weight-bearing joints, systemic processes may be most important in the association between obesity and OA. Although a number of studies assessed individual metabolic factors in relation to hand OA,8 ,17 ,18 the association between the metabolic syndrome and presence of hand OA has not been investigated.

      To gain more insight into the relative contribution of mechanical stress and systemic processes to OA of weight-bearing and non-weight-bearing joints, we examined the association of surrogates for both mechanisms with OA of knees, hands or both. We hypothesised that surrogates for mechanical stress associate predominantly with knee OA, whereas surrogates for systemic processes associate predominantly with presence of hand OA, whether or not co-occurring with knee OA.

      Methods

      Study design and study population

      The Netherlands Epidemiology of Obesity (NEO) study is a population-based prospective cohort study with an oversampling of persons with overweight or obesity, aiming to investigate pathways leading to obesity-related diseases. The present study is a cross-sectional analysis of baseline measurements of the NEO study. Detailed information about the study design and data collection has been described previously.26 In short, men and women between 45 years to 65 years of age with a self-reported BMI ≥27 kg/m2 living in the greater area of Leiden (in the west of The Netherlands) were eligible to participate in the NEO study, resulting in 5002 participants. In addition, all inhabitants aged 45–65 years from one municipality (Leiderdorp) were invited irrespective of their BMI, resulting in 1671 participants (including 605 with BMI ≥27 kg/m2) allowing for a reference distribution of BMI.

      All participants completed questionnaires on demographic and clinical data and visited the NEO study centre for several baseline measurements including extensive physical examination and blood sampling.

      The study was approved by the medical ethics committee of the Leiden University Medical Center and all participants gave written informed consent.

      Measures of body composition

      Measured weight (kg) and height (cm) were used to calculate the BMI (kg/m2). Fat mass (kg) was measured by bioelectrical impedance analysis (BIA) using the Tanita foot-to-foot BIA system TBF-300A Body Composition Analyzer.27 Fat-free mass (kg) was calculated on weight and fat mass. Waist circumference (cm) was measured midway between the lower costal margin and iliac crest with a precision of 0.1 cm.

      Measurement of metabolic factors

      Blood pressure was measured three times, the average was used as diastolic and systolic pressures. Serum concentrations of triglycerides (mmol/L), high density lipoprotein (HDL) cholesterol (mmol/L) and glucose (mmol/L) were measured after an overnight fast.

      The metabolic syndrome was defined according to the Adult Treatment Panel III criteria, based on presence of at least three of the following: (1) elevated waist circumference (men ≥102 cm, women ≥88 cm), (2) elevated triglycerides (≥1.7 mmol/L or drug treatment for elevated triglycerides), (3) reduced HDL cholesterol (men <1.03 mmol/L men, women <1.3 mmol/L or drug treatment for reduced HDL cholesterol), (4) elevated blood pressure (systolic ≥130 mm Hg, diastolic ≥85 mm Hg or antihypertensive medication), (5) elevated fasting glucose (≥5.6 mmol/L or glucose lowering medication).28

      Clinical assessment and OA diagnosis

      Self-reported pain on most days of the prior month and presence of morning stiffness with ≤30 min duration were measured using standardised questions. Physical examination of knees and hands was performed by trained research nurses, using a standardised scoring form. Bony swelling, palpable pain and warmth, crepitus and movement restriction of both knees were assessed. Regarding the hands, bony and soft swellings as well as deformities of the distal interphalangeal, proximal interphalangeal, metacarpophalangeal, carpometacarpal and wrist joints were assessed. Clinical OA was defined according to the American College of Rheumatology clinical criteria.29 ,30 Presence of a knee prosthesis was considered as knee OA.

      Statistical analysis

      Data were analysed using SPSS V.20 and STATA V.12.

      In the NEO study there is an oversampling of persons with a BMI ≥27 kg/m2. To correctly represent associations in the general population,31 adjustments for this oversampling were made by weighting individuals towards the BMI distribution of participants from the Leiderdorp municipality (n=1671),32 whose BMI distribution was similar to the BMI distribution in the general Dutch population.33 All results were based on weighted analyses. Consequently, results apply to a population-based study without oversampling of BMI ≥27 kg/m2.

      Based on OA diagnosis, four groups were defined: (1) individuals without knee or hand OA, (2) individuals with only knee OA, (3) individuals with knee and hand OA, and (4) individuals with only hand OA. We performed logistic regression analyses to examine cross-sectional associations of each of the surrogates for mechanical stress (weight, fat-free mass, fat mass) and systemic processes (fat mass, metabolic syndrome) with each of the three OA types, using individuals without knee or hand OA as reference group. Fat mass probably is a surrogate for mechanical stress and systemic processes. To approximate fat mass as surrogate for mechanical stress, adjustment for metabolic factors was performed, and adjustment for weight was performed to approximate fat mass as surrogate for systemic processes. Associations were expressed as ORs with 95% CIs.

      All continuous variables were standardised by dividing individual values by the SD to be able to compare ORs. Consequently, all ORs describe the risk of OA associated with an increase of 1 SD of the studied variable. All analyses have been adjusted for age, sex, height, smoking, education and ethnicity. Analyses on surrogates for mechanical stress were additionally adjusted for metabolic factors (presence of metabolic syndrome, hypertension, hyperglycaemia, hypertriglyceridaemia and reduced HDL cholesterol) and analyses on surrogates for systemic processes were adjusted for weight.

      Finally, to illustrate the relative importance of mechanical stress and systemic processes, age, sex, height, smoking, education and ethnicity adjusted ORs were calculated for presence of each OA type in three weight categories (based on tertiles of weight of the total study population: <75 kg, 75–90 kg, >90 kg), stratified by the metabolic syndrome. Participants in the lowest weight category without metabolic syndrome served as reference.

      Results

      Population characteristics

      After exclusion of individuals with missing data of BIA (n=31) or physical examination (n=14), data from 6628 participants were analysed. Unweighted baseline characteristics, that is, without taking the oversampling into account, are shown in the online supplementary table. Table 1 shows the weighted baseline characteristics stratified by OA; these characteristics represent the population on which all subsequent results apply. Median (25th–75th centiles) age of the total study population was 56 years (50–61 years), BMI 26 kg/m2 (23–28 kg/m2), 56% women.

      View this table:
      Table 1

      Baseline characteristics of the NEO study population, stratified by OA status

      The prevalence of knee OA alone was 10% (95% CI 9% to 11%), knee and hand OA 4% (95% CI 4% to 5%) and hand OA alone 8% (95% CI 7% to 8%). The prevalence of a knee prosthesis was 1% (95% CI 1% to 1%).

      The percentage of women and median age were higher in individuals with knee, knee and hand, or hand OA as compared with individuals without OA. Furthermore, individuals with knee, hand or knee and hand OA had a higher median fat mass and metabolic syndrome prevalence than individuals without OA.

      Surrogates for mechanical stress and different OA types

      First, we investigated associations of surrogates for mechanical stress with the OA types (table 2).

      View this table:
      Table 2

      Associations of surrogates for mechanical stress with different types of OA (individuals without OA served as reference)

      Weight, fat-free mass and fat mass were positively associated with all OA types. The ORs were highest for presence of knee OA and OA in knee and hand. The ORs per SD weight for example, were 1.55 (95% CI 1.39 to 1.73) for knee OA and 1.52 (95% CI 1.31 to 1.76) for knee and hand OA, meaning that 1 SD of weight (15.95 kg) was associated with a 55% higher odds of having knee OA and a 52% higher odds of having knee and hand OA. The OR for hand OA was 1.25 (95% CI 1.09 to 1.42).

      After additional adjustment for metabolic factors, weight, fat-free mass and fat mass remained associated with knee OA and with OA in knee and hand. The associations with hand OA on the contrary decreased.

      In addition, we assessed if associations between surrogates for mechanical stress and knee OA were stronger for bilateral than for unilateral knee OA. Fully adjusted ORs of weight, fat-free mass and fat mass were higher for bilateral OA (OR 1.68 (95% CI 1.44 to 1.97), 2.29 (95% CI 1.58 to 3.34) and 1.54 (95% CI 1.36 to 1.75), respectively), than for unilateral OA (OR 1.38 (95% CI 1.19 to 1.59), 1.92 (95% CI 1.44 to 2.57) and 1.27 (95% CI 1.12 to 1.44), respectively).

      Surrogates for systemic processes and different OA types

      Next, we assessed associations of surrogates for systemic processes with OA (table 3). Although the OR of fat mass for hand OA was higher (1.17 (95% CI 0.74 to 1.86)) than for knee OA or OA in knee and hand (OR 0.88 (95% CI 0.61 to 1.26) and OR 1.03 (95% CI 0.51 to 2.11), respectively), the associations were not statistically significant after adjustment for weight.

      View this table:
      Table 3

      Associations of surrogates for systemic processes with different types of OA (individuals without OA served as reference)

      The metabolic syndrome, surrogate for systemic processes particularly, was associated with all OA types. However, after additional adjustment for weight the associations with knee OA and knee and hand OA disappeared whereas the metabolic syndrome remained associated with hand OA (OR 1.46 (95% CI 1.06 to 2.02)).

      Relative contribution of weight and metabolic syndrome to different OA types

      Figure 1 illustrates the relative contribution of weight as surrogate for mechanical stress and metabolic syndrome as surrogate for systemic processes to the OA types.

      Figure 1
      Figure 1

      The adjusted ORs and corresponding 95% CIs for osteoarthritis (OA) stratified by weight and metabolic syndrome. Individuals with metabolic Participants in the lowest weight category without metabolic syndrome served as reference. Gray connected lines represent individuals with metabolic syndrome, black connected lines represent individuals without metabolic syndrome. Results are based on weighted analyses of the study population, adjusted for age, sex, smoking, education, ethnicity and height. MetS, metabolic syndrome, representing the odds ratio of the metabolic syndrome for OA, adjusted for the weight catagories.

      The ORs for knee OA were stronger in higher weight categories as compared with the lowest weight category. In addition to the depicted OR representing the highest weight category compared with the lowest in individuals without metabolic syndrome (2.62 (95% CI 1.77 to 3.88)), we calculated the OR of highest versus lowest weight category in individuals with metabolic syndrome (2.30 (95% CI 1.29 to 4.12)). Presence of metabolic syndrome, adjusted for the weight categories, did not result in a higher OR for knee OA (OR 1.16 (95% CI 0.91 to 1.47)). The same was observed in relation to OA in knee and hand.

      In hand OA on the contrary, ORs did not increase with increasing weight (highest vs lowest weight category: OR 1.40 (95% CI 0.89 to 2.21)) in individuals without metabolic syndrome (figure 1) and 0.77 (95% CI 0.39 to 1.51) in individuals with metabolic syndrome. The metabolic syndrome on the other hand was associated with presence of hand OA, adjusted for the weight categories; individuals with metabolic syndrome had a higher OR for hand OA as compared with individuals without metabolic syndrome (OR 1.52 (95% CI 1.10 to 2.09)).

      Discussion

      This study aimed to increase insight into the relative contribution of mechanical stress and systemic processes in the relation between overweight or obesity and OA of weight-bearing and non-weight-bearing joints. Knee OA was associated with surrogates for mechanical stress, adjusted for metabolic factors. Similar results were found for OA in knees and hands. Hand OA was associated with the metabolic syndrome, adjusted for weight.

      A growing body of literature exists on mechanical stress and systemic processes in OA pathogenesis, however innovative in this study is the investigation of the relative contribution of both mechanisms to OA of weight-bearing and non-weight-bearing joints, or of both.

      The association of surrogates for mechanical stress with knee OA has been described previously,20–23 ,34 supporting the concept of excessive mechanical stress on the joint surface of obese individuals resulting in damaged joint tissue. Compression of cartilage might activate mechanoreceptors on chondrocytes, inducing signalling cascades leading to synthesis of inflammatory mediators and tissue remodelling.35 ,36 It is unclear which biomechanical factors are involved in the relation between weight and OA, since a recent study showed that neither a decrease nor increase in knee peak compression force was associated with OA progression.37

      Our finding regarding the metabolic syndrome in relation to knee OA is in accordance with a recent study of Han et al,24 reporting no association between metabolic syndrome and knee OA. Another recent study, by Monira Hussain et al,25 did report an association between metabolic syndrome and knee OA, adjusted for BMI. This discrepancy might be due to differences in OA definition. Where in this study clinical knee OA was assessed following the ACR criteria, Monira Hussain et al defined severe knee OA requiring total knee replacement as OA. Han et al defined knee OA by self-reported physician-made diagnosis. The strength of our study is that knee OA was assessed in all 6628 patients by physical examination. Consequently, OA was diagnosed following the ACR criteria, providing an objective and well validated definition.

      Presence of knee and hand OA has not been investigated before. Our hypothesis was that this type of polyarticular OA might be driven by systemic processes, however we observed no association with surrogates for systemic processes after adjustment for weight. Presence of knee and hand OA was associated with surrogates for mechanical stress, even after adjustment for metabolic factors, like presence of knee OA alone. This observation suggests that co-occurrence of knee and hand OA may not be based on a common underlying pathogenic mechanism, but may represent presence of two different types of OA. Since the association between mechanical stress and knee OA was relatively strong, this association could dominate the association with OA at both sites.

      The association between metabolic syndrome and hand OA, adjusted for weight, has not been reported before. A number of studies assessed individual metabolic factors in relation to OA, however the metabolic syndrome, as composition of different metabolic processes, might act as risk factor beyond the risk of its individual components.38

      The association between metabolic syndrome and hand OA might be explained by systemic inflammation, a main characteristic of the metabolic syndrome. Adipose tissue is known as source of proinflammatory and anti-inflammatory cytokines, which have been related to the metabolic syndrome39 and have been suggested to affect joint tissues.4 ,6 ,7 Visceral fat has been described as the most actively secreting type of adipose tissue,40 and has been associated with the metabolic syndrome.41–45 In addition, in a recently performed study we associated visceral fat with hand OA in men.46

      Strengths of this study are the large study population, extensive characterisation of participants and availability of information of weight-bearing and non-weight-bearing joints.

      However, there are a number of potential limitations. The observed associations were not very strong, however since this study aimed to assess the relative contribution of mechanical stress and systemic processes to the OA types the different ORs provide valuable insight.

      Increased mechanical stress and systemic processes are highly correlated in overweight or obese individuals. Therefore, we adjusted for surrogates for mechanical stress in our analyses on systemic processes and vice versa. Although these analyses identified the relative contribution of both mechanisms for OA, residual confounding may still be present in this observational study. We further minimised residual confounding by adjusting for possible confounders as age, sex and surrogates for socioeconomic status (education, smoking, ethnicity). Unfortunately we did not have information on previous knee injury, which may be a confounder in the associations with knee OA.

      Furthermore, since this is a cross-sectional study, causality is difficult to identify. Since the direction of associations cannot be determined, reverse causation may be present. Longitudinal studies are needed to confirm and further explore associations of mechanical stress and systemic processes with OA.

      Knee and hand OA were diagnosed based on clinical criteria, no X-rays were available. However, the ACR clinical criteria are well validated and have high sensitivity and specificity in diagnosing OA.29 ,30 These criteria include findings at physical examination and presence of symptoms as pain. Since it is known that obese individuals are more likely to report pain,47 they could be more prone to be diagnosed as having OA than non-obese individuals. However, the OA prevalence observed in this study is comparable with the prevalence observed in other population-based studies.18 ,48

      Furthermore, fat mass was measured using foot-to-foot BIA. Although this method has been suggested to overestimate fat mass,49 studies comparing foot-to-foot BIA with hand-to-foot BIA, underwater weighing and dual-energy X-ray absorptiometry reported strong correlations.27 ,50

      We assessed all body composition measures in relation to OA per SD to be able to compare strength of associations observed in this study. It must be noted that the OR of the metabolic syndrome, analysed dichotomously, cannot be directly compared with the ORs of the body composition measures.

      This study suggests that in knee OA, whether or not co-occurring with hand OA, mechanical stress is the most important underlying mechanism, whereas in hand OA alone, systemic processes might contribute most. To gain more insight into the underlying mechanisms, longitudinal research could help to understand how excessive mechanical stress leads to degeneration of joint tissue. In hand OA, the role of the metabolic syndrome might be explored by studying the contribution of the different components of the metabolic syndrome to OA development. Furthermore, future research should focus on the role of systemic inflammation.

      Acknowledgments

      The authors thank all individuals who participated in the Netherlands Epidemiology in Obesity study, and all participating general practitioners for inviting eligible participants. The authors also thank all research nurses for collection of the data and I de Jonge for all data management of the NEO study.

      References

      1. ↵
        1. Anderson JJ,
        2. Felson DT
        . Factors associated with osteoarthritis of the knee in the first national Health and Nutrition Examination Survey (HANES I). Evidence for an association with overweight, race, and physical demands of work. Am J Epidemiol 1988;128:179–89.
        OpenUrlAbstract/FREE Full Text
      2. ↵
        1. Yusuf E,
        2. Nelissen RG,
        3. Ioan-Facsinay A,
        4. et al
        . Association between weight or body mass index and hand osteoarthritis: a systematic review. Ann Rheum Dis 2010;69:761–5.
        OpenUrlAbstract/FREE Full Text
      3. ↵
        1. Brandt KD,
        2. Dieppe P,
        3. Radin E
        . Etiopathogenesis of osteoarthritis. Med Clin North Am 2009;93:1–24, xv.
        OpenUrlCrossRefPubMedWeb of Science
      4. ↵
        1. Dumond H,
        2. Presle N,
        3. Terlain B
        et al. Evidence for a key role of leptin in osteoarthritis. Arthritis Rheum 2003;48:3118–29.
        OpenUrlCrossRefPubMedWeb of Science
      5. ↵
        1. Hu PF,
        2. Bao JP,
        3. Wu LD
        . The emerging role of adipokines in osteoarthritis: a narrative review. Mol Biol Rep 2011;38:873–8.
        OpenUrlCrossRefPubMed
      6. ↵
        1. Pottie P,
        2. Presle N,
        3. Terlain B,
        4. et al
        . Obesity and osteoarthritis: more complex than predicted! Ann Rheum Dis 2006;65:1403–5.
        OpenUrlFREE Full Text
      7. ↵
        1. Tilg H,
        2. Moschen AR
        . Adipocytokines: mediators linking adipose tissue, inflammation and immunity. Nat Rev Immunol 2006;6:772–83.
        OpenUrlCrossRefPubMedWeb of Science
      8. ↵
        1. Dahaghin S,
        2. Bierma-Zeinstra SM,
        3. Koes BW,
        4. et al
        . Do metabolic factors add to the effect of overweight on hand osteoarthritis? The Rotterdam Study. Ann Rheum Dis 2007;66:916–20.
        OpenUrlAbstract/FREE Full Text
      9. ↵
        1. King KB,
        2. Findley TW,
        3. Williams AE,
        4. et al
        . Veterans with diabetes receive arthroplasty more frequently and at a younger age. Clin Orthop Relat Res 2013;471:3049–54.
        OpenUrlCrossRefPubMed
      10. ↵
        1. Schett G,
        2. Kleyer A,
        3. Perricone C
        et al. Diabetes is an independent predictor for severe osteoarthritis: results from a longitudinal cohort study. Diabetes Care 2013;36:403–9.
        OpenUrlAbstract/FREE Full Text
      11. ↵
        1. Nieves-Plaza M,
        2. Castro-Santana LE,
        3. Font YM,
        4. et al
        . Association of hand or knee osteoarthritis with diabetes mellitus in a population of Hispanics from Puerto Rico. J Clin Rheumatol 2013;19:1–6.
        OpenUrlPubMed
      12. ↵
        1. Jungmann PM,
        2. Kraus MS,
        3. Alizai H
        et al. Metabolic risk factors are associated with cartilage degradation assessed by T relaxation time at the knee. Arthritis Care Res (Hoboken) 2013;65:1942–50.
        OpenUrlPubMed
      13. ↵
        1. Kelley KM,
        2. Johnson TR,
        3. Ilan J,
        4. et al
        . Glucose regulation of the IGF response system in chondrocytes: induction of an IGF-I-resistant state. Am J Physiol 1999;276(4 Pt 2):R1164–71.
        OpenUrlWeb of Science
      14. ↵
        1. Rojas-Rodriguez J,
        2. Escobar-Linares LE,
        3. Garcia-Carrasco M,
        4. et al
        . The relationship between the metabolic syndrome and energy-utilization deficit in the pathogenesis of obesity-induced osteoarthritis. Med Hypotheses 2007;69:860–8.
        OpenUrlCrossRefPubMedWeb of Science
      15. ↵
        1. DeGroot J
        . The AGE of the matrix: chemistry, consequence and cure. Curr Opin Pharmacol 2004;4:301–5.
        OpenUrlCrossRefPubMedWeb of Science
      16. ↵
        1. Loeser RF,
        2. Yammani RR,
        3. Carlson CS
        et al. Articular chondrocytes express the receptor for advanced glycation end products: potential role in osteoarthritis. Arthritis Rheum 2005;52:2376–85.
        OpenUrlCrossRefPubMedWeb of Science
      17. ↵
        1. Jonsson H,
        2. Helgadottir GP,
        3. Aspelund T
        et al. Hand osteoarthritis in older women is associated with carotid and coronary atherosclerosis: the AGES Reykjavik study. Ann Rheum Dis 2009;68:1696–700.
        OpenUrlAbstract/FREE Full Text
      18. ↵
        1. Hoeven TA,
        2. Kavousi M,
        3. Clockaerts S
        et al. Association of atherosclerosis with presence and progression of osteoarthritis: the Rotterdam Study. Ann Rheum Dis 2013;72:646–51.
        OpenUrlAbstract/FREE Full Text
      19. ↵
        1. Ghosh P,
        2. Cheras PA
        . Vascular mechanisms in osteoarthritis. Best Pract Res Clin Rheumatol 2001;15:693–709.
        OpenUrlCrossRefPubMed
      20. ↵
        1. Wang Y,
        2. Wluka AE,
        3. English DR
        et al. Body composition and knee cartilage properties in healthy, community-based adults. Ann Rheum Dis 2007;66:1244–8.
        OpenUrlAbstract/FREE Full Text
      21. ↵
        1. Sowers MF,
        2. Yosef M,
        3. Jamadar D,
        4. et al
        . BMI vs body composition and radiographically defined osteoarthritis of the knee in women: a 4-year follow-up study. Osteoarthritis Cartilage 2008;16:367–72.
        OpenUrlCrossRefPubMedWeb of Science
      22. ↵
        1. Abbate LM,
        2. Stevens J,
        3. Schwartz TA,
        4. et al
        . Anthropometric measures, body composition, body fat distribution, and knee osteoarthritis in women. Obesity (Silver Spring) 2006;14:1274–81.
        OpenUrlCrossRefPubMed
      23. ↵
        1. Teichtahl AJ,
        2. Wang Y,
        3. Wluka AE,
        4. et al
        . Obesity and knee osteoarthritis: new insights provided by body composition studies. Obesity (Silver Spring) 2008;16:232–40.
        OpenUrlCrossRefPubMed
      24. ↵
        1. Han CD,
        2. Yang IH,
        3. Lee WS,
        4. et al
        . Correlation between metabolic syndrome and knee osteoarthritis: data from the Korean National Health and Nutrition Examination Survey (KNHANES). BMC Public Health 2013;13:603.
        OpenUrlCrossRefPubMed
      25. ↵
        1. Monira HS,
        2. Wang Y,
        3. Cicuttini FM
        et al. Incidence of total knee and hip replacement for osteoarthritis in relation to the metabolic syndrome and its components: a prospective cohort study. Semin Arthritis Rheum 2014;43:429–36.
        OpenUrlCrossRefPubMed
      26. ↵
        1. de Mutsert R,
        2. den Heijer M,
        3. Rabelink TJ
        et al. The Netherlands Epidemiology of Obesity (NEO) study: study design and data collection. Eur J Epidemiol 2013;28:513–23.
        OpenUrlCrossRefPubMed
      27. ↵
        1. Ritchie JD,
        2. Miller CK,
        3. Smiciklas-Wright H
        . Tanita foot-to-foot bioelectrical impedance analysis system validated in older adults. J Am Diet Assoc 2005;105:1617–19.
        OpenUrlCrossRefPubMedWeb of Science
      28. ↵
        1. Grundy SM,
        2. Cleeman JI,
        3. Daniels SR
        et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005;112:2735–52.
        OpenUrlFREE Full Text
      29. ↵
        1. Altman R,
        2. Alarcon G,
        3. Appelrouth D
        et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hand. Arthritis Rheum 1990;33:1601–10.
        OpenUrlCrossRefPubMedWeb of Science
      30. ↵
        1. Altman R,
        2. Asch E,
        3. Bloch D
        et al. Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association. Arthritis Rheum 1986;29:1039–49.
        OpenUrlCrossRefPubMedWeb of Science
      31. ↵
        1. Korn EL,
        2. Graubard BI
        . Epidemiologic studies utilizing surveys: accounting for the sampling design. Am J Public Health 1991;81:1166–73.
        OpenUrlCrossRefPubMedWeb of Science
      32. ↵
        1. Lumley T
        . Analysis of complex survey samples. 2004. http://www.jstatsoft.org/v09/i08/paper
      33. ↵
        1. Misisterie van VWS
        . Hoeveel mensen hebben overgewicht? 2013. http://www.rivm.nl/nldemaat
      34. ↵
        1. Hochberg MC,
        2. Lethbridge-Cejku M,
        3. Scott WW Jr.
        et al. The association of body weight, body fatness and body fat distribution with osteoarthritis of the knee: data from the Baltimore Longitudinal Study of Aging. J Rheumatol 1995;22:488–93.
        OpenUrlPubMedWeb of Science
      35. ↵
        1. Berenbaum F,
        2. Eymard F,
        3. Houard X
        . Osteoarthritis, inflammation and obesity. Curr Opin Rheumatol 2013;25:114–18.
        OpenUrlCrossRefPubMed
      36. ↵
        1. Ramage L,
        2. Nuki G,
        3. Salter DM
        . Signalling cascades in mechanotransduction: cell-matrix interactions and mechanical loading. Scand J Med Sci Sports 2009;19:457–69.
        OpenUrlCrossRefPubMedWeb of Science
      37. ↵
        1. Henriksen M,
        2. Hunter DJ,
        3. Dam EB
        et al.. Is increased joint loading detrimental to obese patients with knee osteoarthritis? A secondary data analysis from a randomized trial. Osteoarthritis Cartilage 2013;21:1865–75.
        OpenUrlCrossRefPubMedWeb of Science
      38. ↵
        1. Eckel RH,
        2. Grundy SM,
        3. Zimmet PZ
        . The metabolic syndrome. Lancet 2005;365:1415–28.
        OpenUrlCrossRefPubMedWeb of Science
      39. ↵
        1. Mattu HS,
        2. Randeva HS
        . Role of adipokines in cardiovascular disease. J Endocrinol 2013;216:T17–36.
        OpenUrlAbstract/FREE Full Text
      40. ↵
        1. Hamdy O,
        2. Porramatikul S,
        3. Al-Ozairi E
        . Metabolic obesity: the paradox between visceral and subcutaneous fat. Curr Diabetes Rev 2006;2:367–73.
        OpenUrlCrossRefPubMed
      41. ↵
        1. Fox CS,
        2. Massaro JM,
        3. Hoffmann U
        et al. Abdominal visceral and subcutaneous adipose tissue compartments: association with metabolic risk factors in the Framingham Heart Study. Circulation 2007;116:39–48.
        OpenUrlAbstract/FREE Full Text
      42. ↵
        1. Lear SA,
        2. Humphries KH,
        3. Kohli S,
        4. et al
        . Visceral adipose tissue, a potential risk factor for carotid atherosclerosis: results of the Multicultural Community Health Assessment Trial (M-CHAT). Stroke 2007;38:2422–9.
        OpenUrlAbstract/FREE Full Text
      43. ↵
        1. Ditomasso D,
        2. Carnethon MR,
        3. Wright CM,
        4. et al
        . The associations between visceral fat and calcified atherosclerosis are stronger in women than men. Atherosclerosis 2010;208:531–6.
        OpenUrlCrossRefPubMedWeb of Science
      44. ↵
        1. Indulekha K,
        2. Anjana RM,
        3. Surendar J,
        4. et al
        . Association of visceral and subcutaneous fat with glucose intolerance, insulin resistance, adipocytokines and inflammatory markers in Asian Indians (CURES-113). Clin Biochem 2011;44:281–7.
        OpenUrlCrossRefPubMed
      45. ↵
        1. Hanley AJ,
        2. Wagenknecht LE,
        3. Norris JM
        et al. Insulin resistance, beta cell dysfunction and visceral adiposity as predictors of incident diabetes: the Insulin Resistance Atherosclerosis Study (IRAS) Family study. Diabetologia 2009;52:2079–86.
        OpenUrlCrossRefPubMedWeb of Science
      46. ↵
        1. Visser AW,
        2. Ioan-Facsinay A,
        3. de Mutsert R,
        4. et al
        . Adiposity and hand osteoarthritis: the Netherlands Epidemiology of Obesity study. Arthritis Res Ther 2014;16:R19.
        OpenUrlCrossRefPubMed
      47. ↵
        1. Hitt HC,
        2. McMillen RC,
        3. Thornton-Neaves T,
        4. et al
        . Comorbidity of obesity and pain in a general population: results from the Southern Pain Prevalence Study. J Pain 2007;8:430–6.
        OpenUrlCrossRefPubMedWeb of Science
      48. ↵
        1. Haugen IK,
        2. Englund M,
        3. Aliabadi P
        et al. Prevalence, incidence and progression of hand osteoarthritis in the general population: the Framingham Osteoarthritis Study. Ann Rheum Dis 2011;70:1581–6.
        OpenUrlAbstract/FREE Full Text
      49. ↵
        1. Gagnon C,
        2. Menard J,
        3. Bourbonnais A
        et al. Comparison of foot-to-foot and hand-to-foot bioelectrical impedance methods in a population with a wide range of body mass indices. Metab Syndr Relat Disord 2010;8:437–41.
        OpenUrlCrossRefPubMed
      50. ↵
        1. Nunez C,
        2. Gallagher D,
        3. Visser M,
        4. et al
        . Bioimpedance analysis: evaluation of leg-to-leg system based on pressure contact footpad electrodes. Med Sci Sports Exerc 1997;29:524–31.
        OpenUrlCrossRefPubMedWeb of Science

      Supplementary materials

      • Supplementary Data

        This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

        Files in this Data Supplement:

      Footnotes

      • Handling editor Tore K Kvien

      • Collaborators NEO Study Group: Frits R Rosendaal, Renée de Mutsert, Ton J Rabelink, Johannes WA Smit, J Wouter Jukema, Albert de Roos, Saskia le Cessie, Pieter S Hiemstra, Margreet Kloppenburg, Tom WJ Huizinga, Hanno Pijl, Eelco JP de Koning, Willem JJ Assendelft, Pieter H Reitsma, Ko Willems van Dijk, Aiko PJ de Vries, Hildo J Lamb, Ingrid M Jazet, Olaf M Dekkers, Nienke R Biermasz, Jeanet W Blom, Patrick CN Rensen, Christa M Cobbaert (Leiden University Medical Center, Leiden, the Netherlands), Martin den Heijer, Jacqueline M Dekker and Brenda W Penninx (VU Medical Center, Amsterdam, The Netherlands).

      • Contributors AWV performed the statistical analysis, interpreted the data and drafted the manuscript. RdM, SlC, MdH and FRR participated in the study design and revised the manuscript for important intellectual content. MK participated in the study design, was involved in analysing and interpreting the data and in drafting and revising the manuscript. All authors read and approved the final version of the manuscript.

      • Funding The NEO study is supported by the Dutch Arthritis Foundation, the participating Departments, the Division and the Board of Directors of the Leiden University Medical Center, and by the Leiden University, Research Profile Area ‘Vascular and Regenerative Medicine’.

      • Competing interests None.

      • Patient consent Obtained.

      • Ethics approval Leiden University Medical Center.

      • Provenance and peer review Not commissioned; externally peer reviewed.