You are here

Article Text

Article menu

Download PDFPDF

Original article
Factors associated with pain and osteoarthritis at the hip and knee in Great Britain’s Olympians: a cross-sectional study
  1. Dale J Cooper1,
  2. Brigitte E Scammell2,
  3. Mark E Batt3,
  4. Debbie Palmer4
  1. 1 Faculty of Health, Education and Life Sciences, Birmingham City University, Birmingham, UK
  2. 2 Division of Rheumatology, Orthopaedics and Dermatology, School of Medicine, The University of Nottingham, and Arthritis Research UK Centre for Sport, Exercise and Osteoarthritis, Queen’s Medical Centre, Nottingham, UK
  3. 3 Nottingham University Hospitals NHS Trust and Arthritis Research UK Centre for Sport, Exercise and Osteoarthritis, Queen’s Medical Centre, Nottingham, UK
  4. 4 School of Applied Sciences, Edinburgh Napier University, Edinburgh, UK
  1. Correspondence to Dr Dale J Cooper, Faculty of Health, Education and Life Sciences, Birmingham City University, Birmingham, B15 3TN, UK; Dalejcooper{at}hotmail.com

Abstract

Background Knowledge of the epidemiology and potentially modifiable factors associated with musculoskeletal disease is an important first step in injury prevention among elite athletes.

Aim This study investigated the prevalence and factors associated with pain and osteoarthritis (OA) at the hip and knee in Great Britain’s (GB) Olympians aged 40 and older.

Methods This is a cross-sectional study. A survey was distributed to 2742 GB Olympians living in 30 countries. Of the 714 (26.0%) who responded, 605 were eligible for analysis (ie, aged 40 and older).

Results The prevalence of hip and knee pain was 22.4% and 26.1%, and of hip and knee OA was 11.1% and 14.2%, respectively. Using a multivariable model, injury was associated with OA at the hip (adjusted OR (aOR) 10.85; 95% CI 3.80 to 30.96) and knee (aOR 4.92; 95% CI 2.58 to 9.38), and pain at the hip (aOR 5.55; 95% CI 1.83 to 16.86) and knee (aOR 2.65; 95% CI 1.57 to 4.46). Widespread pain was associated with pain at the hip (aOR 7.63; 95% CI 1.84 to 31.72) and knee (aOR 4.77; 95% CI 1.58 to 14.41). Older age, obesity, knee malalignment, comorbidities, hypermobility and weight-bearing exercise were associated with hip and knee OA and/or pain.

Conclusions This study detected an association between several factors and hip and knee pain/OA in retired GB Olympic athletes. These associations require further substantiation in retired athletes from other National Olympic Committees, and through comparison with the general population. Longitudinal follow-up is needed to investigate the factors associated with the onset and progression of OA/pain, and to determine if modulation of such factors can reduce the prevalence of pain and OA in this population.

  • hip
  • knee
  • osteoarthritis
  • post olympic
  • health

http://dx.doi.org/10.1136/bjsports-2017-098315

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.

    Background

    A key priority of the IOC and its Medical Commission is to protect the health of the athletes in sport.1 During recent years, the IOC has promoted research to prevent injuries and illnesses in sport by determining injury epidemiology, risk factors, injury mechanisms and interventions to protect athletes’ health. Yet the long-term musculoskeletal health of the athlete has received far less attention. Data from retired athletes are a valuable source of information for a number of reasons. First, it is important to understand the diseases affecting retired athletes in order to determine if there is a need for prevention. Second, data from retired athletes can help to determine if there are modifiable risk factors that can protect the long-term health of athletes.

    Musculoskeletal diseases such as pain and osteoarthritis (OA) are likely to adversely impair a retired athlete’s quality of life—morbidity associated with knee OA is high,2 and years lived with disability for knee OA are substantial.3 Previous studies have found that, compared with the general population, retired male elite athletes are at an increased risk of developing OA.4–6 However, putative risk factors associated with pain and OA in non-sporting populations remain substantially unexplored in retired elite athletes. Therefore, in view of the responsibility to protect the long-term health of all athletes, it is essential to identify the risk factors that associate with musculoskeletal disease in later life. This study aimed to determine in Great Britain’s (GB) Olympians aged 40 years and older (1) the prevalence of pain and OA at the hip and knee, and (2) the factors that are associated with pain and OA at the hip and knee.

    Methods

    Study design

    This study was cross-sectional and involved distributing a survey to collect information on factors potentially associated with pain and OA at the hip and knee, as well as demographics, medical history, drug history, general health, and occupational history including participation in sport and physical activity. Implied consent to participate was obtained from all participants completing the study questionnaire.

    Eligibility criteria and setting

    Recruitment took place between May 2014 and April 2015. Initial contact was made by placing an advertisement for the study in the British Olympic Association (BOA) membership magazine. The BOA Athletes’ Commission then distributed a letter by post or email, inviting GB Olympians listed on the BOA Olympian database and giving them the opportunity to complete and return a paper or web-based version of the questionnaire. One reminder was sent by post to those who did not respond within 4 weeks. The inclusion criteria for participants were male or female, aged 40 years and older, and (1) must have represented GB at the Summer and/or the Winter Olympic Games; (2) were registered on the BOA Olympian database; and (3) were able to give informed consent.

    Data collection and management

    The design of the questionnaire was based on two previously published questionnaires7 8 and was available in two formats: (1) a paper-based version and (2) a web-based version hosted by Bristol University Survey. The content and clarity of the questionnaire were reviewed in a patient public involvement (PPI) focus group interview with local residents (n=6) and the Committee at the BOA Athletes’ Commission (n=14). The questionnaire was assessed as part of two pilot studies at the research institution (n=12). All amendments were returned to the PPI members for verification.

    The questionnaire was designed to collect detailed information including age (years), sex, ethnicity, body mass index (BMI, kg/m2), and putative risk factors associated with pain and OA at the hip and knee. The questionnaire incorporated a validated screening question7 that was also adapted for hip pain: ‘have you ever had knee pain for most days of the past one month?’ A body manikin was used as a self-report screening instrument to record the location of hip and knee pain and pain in other body regions, using a method shown to be repeatable.9 Chronic widespread pain was recorded if an individual had greater than or equal to 7 out of 19 regions on the Widespread Pain Index.10 The presence of OA was determined by asking participants ‘have you ever been diagnosed with osteoarthritis in any of your joints by a physician, and if so, please state which joint/s?’ The presence of finger nodes and the index to ring finger ratio (2D:4D) were determined using validated diagrams.7 11 12 Finger nodes were classified as present in those self-reporting nodal changes on at least two rays of both hands. The visual classification of the index to ring finger ratio consisted of classifying each hand according to whether the index finger was visually longer than (type 1), equal to (type 2) or shorter than (type 3) the ring finger. Joint flexibility was determined by self-examination using line drawings of nine genetically determined sites from the 9-point Beighton score.13 A cut-off threshold of equal to or greater than 4 out of 9 on the modified Beighton 9-point scoring system was used to denote generalised joint hypermobility, as recommended by the British Society of Rheumatology.14 Knee alignment was assessed using a validated line drawing instrument.15 Knee alignment grades were classified according to the following: A=severe varus, B=mild varus, C=straight legs, D=mild valgus and E=severe valgus. Early life (ie, during the 20s) and current measures of joint flexibility and knee alignment were recorded separately. The questionnaire captured information on comorbidities (ie, diabetes, cancer, lung disease, stroke, heart disease), and previous significant injuries and surgery. Comorbidities were graded into (1) those who were not reported to be suffering from one or more comorbidities, (2) those suffering from a single comorbidity and (3) those suffering from two or more. The presence of a significant injury was determined by asking participants ‘have you ever sustained a significant injury that caused pain for most days during a one month period and for which you consulted a medical professional or a health provider such as a general practitioner?’ The sporting discipline in which participants competed in at the Olympic Games was categorised into impact sports and non-impact sports, and weight-bearing and non-weight-bearing sports, based on published evidence.4 16 Where GB Olympians had competed in at least two disciplines at the Olympic level, preference was given to the discipline in which the participant had spent the longest time competing.

    Statistical analysis

    Questionnaire data were entered into an Excel file. Data were then cleaned, coded and analysed using SPSS V.22.0. The prevalence of the primary outcome variables of pain and OA was calculated using the most severe hip or knee joint. Crude ORs with 95% CIs were computed using logistic regression to determine the univariate associations between each independent variable and the outcome variables. Age and BMI were non-linear and categorised according to previous research.8 Significant injuries were included if they were reported to have preceded the date of diagnosis of OA or episode of pain. All significant factors at p<0.05 were entered separately into a second model and adjusted for a priori confounders of age, sex and BMI.8 A mutually adjusted model was then fitted of the a priori confounders plus any significant factors/variables. A final check was undertaken to refit, one at a time, the independent variables excluded from earlier models. Imputation was not undertaken for the occasional missing values.

    Study power

    A power calculation was based on the assumption of approximately 14% and 19% prevalence of hip and knee OA,4 and 15% and 25% for hip and knee pain.16 17 With the assumptions of a 30% response rate from GB Olympians aged 40 years and older, assuming all exposures could at least be dichotomised into binary variables and assuming a ratio of exposed to unexposed individuals of 1:1 for any given factor, the study had power of at least 80% to detect ORs of 1.75 and 1.85 or greater for knee pain and knee OA, respectively, at 5% significance. Similarly this applies to hip pain and hip OA for an OR of 2.0 or greater.

    Results

    Characteristics of the participants

    The overall response rate to the questionnaire was 26.0% (714/2742). Of those who replied to the questionnaire, 605 were 40 years or older and had data for the analysis. This represents 32.1% (605/1887) of the cohort on the BOA Olympian database who were aged 40 and older in 2015 (see figure 1). Of those included in the analysis, the mean age was 63.6±13.3 years, 59.7% were male (361/605) and 40.3% were female (244/605) (see table 1). Of the 605 respondents, 60 had competed in 11 sports at the Winter Olympic Games: alpine skiing (12), bobsleigh (12), figure skating (10), cross-country skiing (9), luge (4), biathlon (4), short track speed skating (4), speed skating (2), ice hockey (1), skeleton (1) and freestyle skiing (1); and 545 had competed in 25 sporting disciplines at the Summer Olympic Games: athletics (144), rowing (87), swimming (65), hockey (51), canoeing (27), cycling (25), fencing (22), gymnastics (20), sailing (18), archery (11), equestrian (11), shooting (10), diving (10), judo (8), boxing (7), weightlifting (7), football (5), wrestling (3), basketball (3), water polo (3), tennis (2), badminton (2), synchronised swimming (2), table tennis (1) and windsurfing (1).

    View this table:
    Table 1

    Anthropometry, lifestyle and health factors

    Figure 1
    Figure 1

    Flow chart describing the number of retired Olympic athletes included in this study from the British Olympic Association database in 2015. The flow chart describes those who could not be contacted, the number of surveys distributed, the number of surveys returned and the number of surveys included in the analysis meeting the inclusion criteria.

    Prevalence of pain and OA

    The prevalence of hip and knee pain was 22.4% (126/563) and 26.1% (147/564), and of hip and knee OA was 11.1% (66/597) and 14.2% (85/597), respectively. The results of the multivariable regression models are presented in tables 2–5.

    View this table:
    Table 2

    Constitutional/biomechanical factors and prevalence of knee pain (n=564)

    View this table:
    Table 3

    Constitutional/biomechanical factors and prevalence of knee osteoarthritis (n=597)

    View this table:
    Table 4

    Constitutional/biomechanical factors and prevalence of hip pain (n=563)

    View this table:
    Table 5

    Constitutional/biomechanical factors and prevalence of hip osteoarthritis (n=597)

    Factors associated with knee pain and knee OA

    Knee pain was associated with widespread pain (adjusted OR (aOR) 4.77; 95% CI 1.58 to 14.41, p=0.006), obesity (kg/m2) (aOR 4.34; 95% CI 2.30 to 8.19, p<0.001), knee injury (aOR 2.65; 95% CI 1.57 to 4.46, p<0.001) and older age (aOR 1.61; 95% CI 1.02 to 2.53, p=0.04). There was some evidence that participation in weight-bearing sport (aOR 1.61; 95% CI 1.06 to 2.44, p=0.027) was associated with knee pain only if adjusted for age, sex and BMI (see table 2). Knee OA was associated with knee injury (aOR 4.92; 95% CI 2.58 to 9.38, p<0.001), older age (aOR 3.49; 95% CI 1.71 to 7.11, p=0.001), early-life (ie, during the 20s) varus knee malalignment (aOR 2.97; 95% CI 1.11 to 7.94, p=0.03), early-life joint hypermobility (aOR 2.64; 95% CI 1.21 to 5.78, p=0.015), comorbidities (two or more) (aOR 2.61; 95% CI 1.23 to 5.52, p=0.012) and obesity (kg/m2) (aOR 2.35; 95% CI 1.03 to 5.38, p=0.042) (see table 3).

    Factors associated with hip pain and hip OA

    Hip pain was associated with prior injury (aOR 5.55; 95% CI 1.83 to 16.86, p=0.002), widespread pain (aOR 7.63; 95% CI 1.84 to 31.72, p=0.005), participation in weight-bearing sport (aOR 1.66; 95% CI 1.05 to 2.63, p=0.032) and comorbidities (aOR 1.84; 95% CI 1.05 to 3.22, p=0.033) (see table 4). Hip OA was also associated with prior hip injury (aOR 10.85; 95% CI 3.80 to 30.96, p<0.001), older age (aOR 2.93; 95% CI 1.48 to 5.82, p=0.002) and comorbidities (aOR 2.46; 95% CI 1.19 to 5.06, p=0.015) (see table 5).

    Discussion

    This study investigated the prevalence and factors associated with pain and OA at the hip and knee in GB Olympians aged 40 and older. The present study found that (1) pain at the hip (22.4%; 126/563) and knee (26.1%; 147/564), as well as OA at the hip (11.1%; 66/597) and knee (14.2%; 85/597), are prevalent disorders in GB Olympians aged 40 and older; (2) significant injury was associated with hip and knee OA, and pain at the hip and knee; (3) bodily pain at other sites (ie, widespread pain) was associated with hip and knee pain; (4) early-life knee malalignment and joint hypermobility (self-report Beighton ≥4/9) were not associated with pain and OA, with the exception of knee OA; (5) retired athletes with two or more comorbidities were more likely to report hip pain, and hip and knee OA; and (6) participation in impact (ie, contact) sport was not associated with pain and OA. It remains unclear if participation in weight-bearing sports is associated with future hip and knee pain or OA.

    Comparisons with other studies

    The paucity of existing data limits the number of comparisons that can be made with other sporting populations. The present study found the prevalence of knee pain of 26.1% is similar, although slightly higher, than that previously found in non-sporting community populations,18 19 but lower compared with that found in retired male international athletes who had competed in fewer Olympic sporting disciplines.5 The present study found a higher prevalence of hip and knee OA of 11.1% and 14.2%, compared with previous observations in community populations.20 21 Yet the prevalence of OA at the hip and knee was lower than that found in 709 former internationally or nationally ranked Swedish athletes,4 and 991 male former athletes who had represented Finland in international competitions,5 using an identical self-report, physician-diagnosed definition of OA. Direct comparisons with other cohort studies including the general population are problematic, mainly due to the different age distribution of the study participants, different case definitions and variations in how prevalence is calculated between studies.

    The present study found a higher prevalence of knee OA and pain at the hip and knee in GB Olympians aged 60 and older compared with those aged 40–59. Previous studies in the general population confirm that older age is a constitutional risk factor for OA at the hip22 and knee,23–26 as well as knee pain.27 There was also a significant association between obesity and pain/OA at the knee. This is consistent with findings from previous cohort studies of knee OA28 29 and knee pain.8 30 Obesity is commonly believed to affect joints through biomechanical loading, although more recent studies provide evidence of a metabolic inflammatory pathway between BMI and knee OA.31 32

    Previous observations in the general population posit injury as a major risk factor for the development of knee OA33 34 and knee pain.8 27 35 The present study confirmed a significant association between injury and hip and knee OA/pain. Meniscal injuries, dislocations, fractures36 and ACL tears37 38 have all been shown to increase the risk of knee OA. Direct trauma to the tissue may disrupt normal joint kinematics and cause altered load distribution within the joint, and this is thought to contribute to the initiation of OA.36 For the present study, all the knee cartilage injuries sustained in competition or training among GB Olympians occurred during weight-bearing activities.

    Long-term weight-bearing sports activity was associated with a twofold to threefold increase in the risk of radiographic hip and knee OA in middle-aged ex-elite athletes and a subgroup of the general population who reported long-term sports activity.16 The present study found an association between participation in weight-bearing sport and hip pain, and knee pain, but only if adjusted for age, sex and BMI. It remains unclear if participation in weight-bearing sports is associated with future hip and knee pain or OA. Furthermore, participation in impact (ie, contact) sport was not detected to be associated with hip and knee pain or OA. A previous study4 reported that retired male athletes who participated in impact sports at an elite level had an increased prevalence of self-report, physician-diagnosed knee OA following adjustment for age, BMI and occupational load. However, this increased risk from participating in impact sports was within a population consisting largely of ex-professional football players, and was driven by an increased risk of joint injury. The present study population included retired athletes from a wide range of sporting disciplines.

    Knee malalignment is thought to contribute to cartilage degeneration through an alteration in the load distribution acting across the articular surfaces of the tibiofemoral joint.36 A case–control study of 1901 patients found early-life knee malalignment (especially varus) was associated with the later development of knee OA.39 The same self-reported instrument was also used in a cohort study of 2156 healthy controls and found that early-life self-reported knee varus or valgus malalignment was also a cause of knee pain.8 Used in this study, the same self-reported instrument confirmed that early-life varus knee malalignment is associated with knee OA in retired elite athletes. This study found no association between knee varus malalignment and knee pain, nor did this study detect an association between valgus malalignment and knee pain or knee OA. These findings are consistent with previous studies that tend to show more positive associations between varus knee malalignment and the development of knee OA in the general population.39 40

    Hypermobile joints are thought to exert greater biomechanical stresses on articular cartilage, and this may increase the risk of OA and pain. Although a correlation between joint hypermobility and OA appears to be possible in community populations,41–44 there is a lack of evidence to conclude whether joint hypermobility acts as a risk factor or as a protector from the development of pain and OA. In the present cohort, there was no association detected between self-report joint hypermobility in early life with the various outcomes other than knee OA. Those suffering from two or more comorbidities (ie, diabetes, cancer, lung disease, stroke, heart disease) were more likely to report hip and knee OA, as well as hip pain. This study did not detect an association between the index to ring finger length (2D:4D) ratio and knee OA. This was in contrast to a previous study12 that demonstrated that individuals in the general population with male patterning (ie, type 3—index finger shorter than ring finger) were at greater risk of knee OA than those with a different finger patterning. This lack of association is possibly due to the present study using a self-report instrument compared with a radiographic measurement used in the previous study to determine the index to ring finger ratio.

    In community populations, bodily pain at other sites (ie, widespread pain) was associated with knee pain in knee OA.30 45 The present study found hip and knee pain to be more prevalent than hip and knee OA, respectively, and that an association existed between widespread pain and pain at the hip and knee. The findings of this study suggest that a subset of GB Olympians may have a chronic widespread pain disorder, and that persistent hip and knee pain in those aged 40 and older is not a surrogate of self-reported, physician-diagnosed hip and knee OA.

    Strengths and limitations of this study

    The strengths of the present study include a large population sample with a wide age range from both sexes. The analysis represents approximately a third of all GB Olympians aged 40 years and older in 2015. This study used validated patient-reported outcome measures. The findings of the present study concur with previous studies in cohorts of non-sporting elite athletes: studies indicating that age, obesity and previous joint injury are associated with pain and OA. This study also detected that age, comorbidities, widespread pain, weight-bearing sports, and early-life knee malalignment and joint hypermobility were associated with the prevalence of pain and/or OA in retired elite athletes. However, this study did not find any association with participation in an impact sport, length of the index to ring finger ratio, finger nodes and sex. Thus, this study extends previous findings and contributes to the knowledge of factors associated with pain and OA in retired elite athletes.

    This study was not without its limitations. First, the results of this study (eg, history of injury/OA/ joint hypermobility) are subject to potential recall bias. Second, the use of BMI was potentially misleading; triceps skinfold thickness (peripheral fat) in men and the waist to hip ratio (central fat) in women are demonstrated to be more strongly associated with knee OA than BMI.46 Furthermore, BMI is unable to discriminate between muscle and adipose tissue, which may be particularly pertinent in a retired elite sporting population, and it cannot directly assess regional adiposity.47 Third, one should apply caution when assuming that there is a direct causality between factors and the outcome, as other explanations may exist, and this study cannot exclude the possibility of residual confounding. The cross-sectional design is subject to limitations of temporality and future cohort studies can better demonstrate that causes preceded the outcome. Fourth, internal validity was increased through the use of internal controls, although this reduced the generalisability of the findings to the general population and retired athletes from other National Olympic Committees as the sports included reflect those Olympic events most pursued by GB. Fifth, despite the strenuous efforts to achieve a high response rate—all GB Olympians on the BOA Olympian database were invited to participate in this study—there is a possibility of recruitment bias. Sixth, the crude OR for hip injury and OA is large and mildly inflated in multivariable analyses, and this may reflect sparse-data bias as a result of the small number of cases of hip injury and OA.48 49 Penalisation was not undertaken as the events per covariate were above 5.48

    Conclusions

    This study reports early important work on the long-term musculoskeletal health of retired Olympic athletes. This study detected an association between several factors and hip and knee pain and/or OA in retired GB Olympians. These associations require further substantiation in retired athletes from other National Olympic Committees and through comparison with the general population. Longitudinal follow-up is needed to investigate the onset and progression of OA/pain, and to determine if modulation of such factors can reduce the prevalence of pain and OA in this population. Strategies to treat one of the mechanisms of pain for all retired athletes may have low efficacy, should the pain in some retired athletes be mediated by other mechanisms. Further research is required to identify the factors associated with different pain mechanisms in non-sporting and sporting populations including retired athletes from other National Olympic Committees.

    What are the findings?

    • Significant joint injury was strongly associated with self-reported hip and knee osteoarthritis (OA), and hip and knee pain.

    • Bodily pain at other sites (ie, widespread pain) was strongly associated with self-reported hip and knee pain.

    • Participation in impact (ie, contact) sport was not associated with hip and knee pain or OA.

    • It remains unclear if participation in weight-bearing sports is associated with future hip and knee pain or OA.

    How might it impact on clinical practice in the future?

    • The evidence demonstrates an association between significant joint injury and pain/OA at the hip and knee in retired Great Britain’s Olympic athletes.

    • Further research should focus on the factors for joint injury in different sports.

    • This information may help to develop interventions to protect the long-term health of athletes.

    • Medical staff, athletes, coaches and key stakeholders should seek to integrate injury prevention programmes in daily training to minimise the long-term health risks associated with joint injury.

    Acknowledgments

    We would like to thank the staff at the Arthritis Research UK Centre for Sport, Exercise and Osteoarthritis, The University of Nottingham, the British Olympic Association (BOA) Athletes’ Commission, and all the participants who contributed to this study. We would like to thank Sarah Winckless and Christine Bower for their invaluable assistance with distributing the survey.

    References

    1. ↵
      2. Engebretsen L ,
      3. Bahr R ,
      4. Cook JL , et al
      . The IOC Centres of Excellence bring prevention to sports medicine. Br J Sports Med 2014;48:1270–5.doi:10.1136/bjsports-2014-093992
      OpenUrlAbstract/FREE Full Text
    2. ↵
      2. Vos T ,
      3. Flaxman AD ,
      4. Naghavi M , et al
      . Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2163–96.doi:10.1016/S0140-6736(12)61729-2
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      2. Cross M ,
      3. Smith E ,
      4. Hoy D , et al
      . The global burden of hip and knee osteoarthritis: estimates from the global burden of disease 2010 study. Ann Rheum Dis 2014;73:1323–30.doi:10.1136/annrheumdis-2013-204763
      OpenUrlAbstract/FREE Full Text
    4. ↵
      2. Tveit M ,
      3. Rosengren BE ,
      4. Nilsson JÅ , et al
      . Former male elite athletes have a higher prevalence of osteoarthritis and arthroplasty in the hip and knee than expected. Am J Sports Med 2012;40:527–33.doi:10.1177/0363546511429278
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      2. Kettunen JA ,
      3. Kujala UM ,
      4. Kaprio J , et al
      . Lower-limb function among former elite male athletes. Am J Sports Med 2001;29:2–8.doi:10.1177/03635465010290010801
      OpenUrlPubMedWeb of Science
    6. ↵
      2. Kujala UM ,
      3. Sarna S ,
      4. Kaprio J , et al
      . Heart attacks and lower-limb function in master endurance athletes. Med Sci Sports Exerc 1999;31:1041–6.doi:10.1097/00005768-199907000-00018
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      2. O’Reilly S ,
      3. Johnson S ,
      4. Doherty S , et al
      . Screening for hand osteoarthritis (OA) using a postal survey. Osteoarthritis Cartilage 1999;7:461–5.doi:10.1053/joca.1999.0240
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      2. Ingham SL ,
      3. Zhang W ,
      4. Doherty SA , et al
      . Incident knee pain in the Nottingham community: a 12-year retrospective cohort study. Osteoarthritis Cartilage 2011;19:847–52.doi:10.1016/j.joca.2011.03.012
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      2. Lacey RJ ,
      3. Lewis M ,
      4. Jordan K , et al
      . Interrater reliability of scoring of pain drawings in a self-report health survey. Spine 2005;30:E455–E458.doi:10.1097/01.brs.0000174274.38485.ee
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      2. Wolfe F ,
      3. Clauw DJ ,
      4. Fitzcharles MA , et al
      . The American College of Rheumatology preliminary diagnostic criteria for fibromyalgia and measurement of symptom severity. Arthritis Care Res 2010;62:600–10.doi:10.1002/acr.20140
      OpenUrlCrossRefWeb of Science
    11. ↵
      2. Rees F ,
      3. Doherty S ,
      4. Hui M , et al
      . Distribution of finger nodes and their association with underlying radiographic features of osteoarthritis. Arthritis Care Res 2012;64:533–8.doi:10.1002/acr.21586
      OpenUrl
    12. ↵
      2. Zhang W ,
      3. Robertson J ,
      4. Doherty S , et al
      . Index to ring finger length ratio and the risk of osteoarthritis. Arthritis Rheum 2008;58:137–44.doi:10.1002/art.23237
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      2. Cooper DJ ,
      3. Scammell BE ,
      4. Batt ME , et al
      . Development and validation of self-reported line drawings of the modified Beighton score for the assessment of generalised joint hypermobility. BMC Med Res Methodol 2018;18:11.doi:10.1186/s12874-017-0464-8
      OpenUrl
    14. ↵
      2. Remvig L ,
      3. Jensen DV ,
      4. Ward RC
      . Are diagnostic criteria for general joint hypermobility and benign joint hypermobility syndrome based on reproducible and valid tests? A review of the literature. J Rheumatol 2007;34:798–803.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      2. Ingham SL ,
      3. Moody A ,
      4. Abhishek A , et al
      . Development and validation of self-reported line drawings for assessment of knee malalignment and foot rotation: a cross-sectional comparative study. BMC Med Res Methodol 2010;10:57.doi:10.1186/1471-2288-10-57
      OpenUrlPubMed
    16. ↵
      2. Spector TD ,
      3. Harris PA ,
      4. Hart DJ , et al
      . Risk of osteoarthritis associated with long-term weight-bearing sports: a radiologic survey of the hips and knees in female ex-athletes and population controls. Arthritis Rheum 1996;39:988–95.doi:10.1002/art.1780390616
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      2. Peat G ,
      3. McCarney R ,
      4. Croft P
      . Knee pain and osteoarthritis in older adults: a review of community burden and current use of primary health care. Ann Rheum Dis 2001;60:91–7.doi:10.1136/ard.60.2.91
      OpenUrlAbstract/FREE Full Text
    18. ↵
      2. Turkiewicz A ,
      3. Gerhardsson de Verdier M ,
      4. Engström G , et al
      . Prevalence of knee pain and knee OA in southern Sweden and the proportion that seeks medical care. Rheumatology 2015;54:827–35.doi:10.1093/rheumatology/keu409
      OpenUrlCrossRefPubMed
    19. ↵
      2. Cecchi F ,
      3. Mannoni A ,
      4. Molino-Lova R , et al
      . Epidemiology of hip and knee pain in a community based sample of Italian persons aged 65 and older. Osteoarthritis Cartilage 2008;16:1039–46.doi:10.1016/j.joca.2008.01.008
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      2. Plotnikoff R ,
      3. Karunamuni N ,
      4. Lytvyak E , et al
      . Osteoarthritis prevalence and modifiable factors: a population study. BMC Public Health 2015;15:1–10.doi:10.1186/s12889-015-2529-0
      OpenUrlCrossRefPubMed
    21. ↵
      2. Grotle M ,
      3. Hagen KB ,
      4. Natvig B , et al
      . Obesity and osteoarthritis in knee, hip and/or hand: an epidemiological study in the general population with 10 years follow-up. BMC Musculoskelet Disord 2008;9:132.doi:10.1186/1471-2474-9-132
      OpenUrlCrossRefPubMed
    22. ↵
      2. Dagenais S ,
      3. Garbedian S ,
      4. Wai EK
      . Systematic review of the prevalence of radiographic primary hip osteoarthritis. Clin Orthop Relat Res 2009;467:623–37.doi:10.1007/s11999-008-0625-5
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      2. Blagojevic M ,
      3. Jinks C ,
      4. Jeffery A , et al
      . Risk factors for onset of osteoarthritis of the knee in older adults: a systematic review and meta-analysis. Osteoarthritis Cartilage 2010;18:24–33.doi:10.1016/j.joca.2009.08.010
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      2. Toivanen AT ,
      3. Heliövaara M ,
      4. Impivaara O , et al
      . Obesity, physically demanding work and traumatic knee injury are major risk factors for knee osteoarthritis--a population-based study with a follow-up of 22 years. Rheumatology 2010;49:308–14.doi:10.1093/rheumatology/kep388
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      2. Jordan JM ,
      3. Helmick CG ,
      4. Renner JB , et al
      . Prevalence of knee symptoms and radiographic and symptomatic knee osteoarthritis in African Americans and Caucasians: the Johnston County Osteoarthritis Project. J Rheumatol 2007;34:172–80.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      2. Hochberg MC ,
      3. Lethbridge-Cejku M ,
      4. Tobin JD
      . Bone mineral density and osteoarthritis: data from the Baltimore Longitudinal Study of Aging. Osteoarthritis Cartilage 2004;12 Suppl A:S45–8.
      OpenUrlCrossRefPubMed
    27. ↵
      2. Miranda H ,
      3. Viikari-Juntura E ,
      4. Martikainen R , et al
      . A prospective study on knee pain and its risk factors. Osteoarthritis Cartilage 2002;10:623–30.doi:10.1053/joca.2002.0796
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      2. Cooper C ,
      3. Snow S ,
      4. McAlindon TE , et al
      . Risk factors for the incidence and progression of radiographic knee osteoarthritis. Arthritis & Rheumatism 2000;43:995–1000.doi:10.1002/1529-0131(200005)43:5<995::AID-ANR6>3.0.CO;2-1
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      2. Hochberg MC ,
      3. Lethbridge-Cejku M ,
      4. Scott WW , 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
    30. ↵
      2. Jinks C ,
      3. Jordan KP ,
      4. Blagojevic M , et al
      . Predictors of onset and progression of knee pain in adults living in the community. A prospective study. Rheumatology 2008;47:368–74.doi:10.1093/rheumatology/kem374
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      2. Karvonen-Gutierrez CA ,
      3. Harlow SD ,
      4. Jacobson J , et al
      . The relationship between longitudinal serum leptin measures and measures of magnetic resonance imaging-assessed knee joint damage in a population of mid-life women. Ann Rheum Dis 2014;73:883–9.doi:10.1136/annrheumdis-2012-202685
      OpenUrlAbstract/FREE Full Text
    32. ↵
      2. Fowler-Brown A ,
      3. Kim DH ,
      4. Shi L , et al
      . The mediating effect of leptin on the relationship between body weight and knee osteoarthritis in older adults. Arthritis Rheumatol 2015;67:169–75.doi:10.1002/art.38913
      OpenUrl
    33. ↵
      2. Muthuri SG ,
      3. McWilliams DF ,
      4. Doherty M , et al
      . History of knee injuries and knee osteoarthritis: a meta-analysis of observational studies. Osteoarthritis Cartilage 2011;19:1286–93.doi:10.1016/j.joca.2011.07.015
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      2. Wilder FV ,
      3. Hall BJ ,
      4. Barrett JP , et al
      . History of acute knee injury and osteoarthritis of the knee: a prospective epidemiological assessment. The Clearwater Osteoarthritis Study. Osteoarthritis Cartilage 2002;10:611–6.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      2. Silverwood V ,
      3. Blagojevic-Bucknall M ,
      4. Jinks C , et al
      . Current evidence on risk factors for knee osteoarthritis in older adults: a systematic review and meta-analysis. Osteoarthritis Cartilage 2015;23:507–15.doi:10.1016/j.joca.2014.11.019
      OpenUrlCrossRefPubMed
    36. ↵
      2. Litwic A ,
      3. Edwards MH ,
      4. Dennison EM , et al
      . Epidemiology and burden of osteoarthritis. Br Med Bull 2013;105:185–99.doi:10.1093/bmb/lds038
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      2. Harkey MS ,
      3. Luc BA ,
      4. Golightly YM , et al
      . Osteoarthritis-related biomarkers following anterior cruciate ligament injury and reconstruction: a systematic review. Osteoarthritis Cartilage 2015;23:1–12.doi:10.1016/j.joca.2014.09.004
      OpenUrlCrossRefPubMed
    38. ↵
      2. Ajuied A ,
      3. Wong F ,
      4. Smith C , et al
      . Anterior cruciate ligament injury and radiologic progression of knee osteoarthritis: a systematic review and meta-analysis. Am J Sports Med 2014;42:2242–52.doi:10.1177/0363546513508376
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      2. McWilliams DF ,
      3. Doherty S ,
      4. Maciewicz RA , et al
      . Self-reported knee and foot alignments in early adult life and risk of osteoarthritis. Arthritis Care Res 2010;62:489–95.doi:10.1002/acr.20169
      OpenUrl
    40. ↵
      2. Brouwer GM ,
      3. van Tol AW ,
      4. Bergink AP , et al
      . Association between valgus and varus alignment and the development and progression of radiographic osteoarthritis of the knee. Arthritis Rheum 2007;56:1204–11.doi:10.1002/art.22515
      OpenUrlCrossRefPubMedWeb of Science
    41. ↵
      2. Jónsson H ,
      3. Elíasson GJ ,
      4. Jónsson A , et al
      . High hand joint mobility is associated with radiological CMC1 osteoarthritis: the AGES-Reykjavik study. Osteoarthritis Cartilage 2009;17:592–5.doi:10.1016/j.joca.2008.10.002
      OpenUrlCrossRefPubMedWeb of Science
    42. ↵
      2. Scott D ,
      3. Bird H ,
      4. Wright V
      . Joint laxity leading to osteoarthrosis. Rheumatol Rehabil 1979;18:167–9.doi:10.1093/rheumatology/18.3.167
      OpenUrlCrossRefPubMedWeb of Science
    43. ↵
      2. Jónsson H ,
      3. Valtýsdóttir ST ,
      4. Kjartansson O , et al
      . Hypermobility associated with osteoarthritis of the thumb base: a clinical and radiological subset of hand osteoarthritis. Ann Rheum Dis 1996;55:540–3.doi:10.1136/ard.55.8.540
      OpenUrlAbstract/FREE Full Text
    44. ↵
      2. Jónsson H ,
      3. Valtýsdóttir ST
      . Hypermobility features in patients with hand osteoarthritis. Osteoarthritis Cartilage 1995;3:1–5.doi:10.1016/S1063-4584(05)80032-9
      OpenUrlPubMed
    45. ↵
      2. Croft P ,
      3. Jordan K ,
      4. Jinks C
      . “Pain elsewhere” and the impact of knee pain in older people. Arthritis & Rheumatism 2005;52:2350–4.doi:10.1002/art.21218
      OpenUrlCrossRefPubMedWeb of Science
    46. ↵
      2. Sanghi D ,
      3. Srivastava RN ,
      4. Singh A , et al
      . The association of anthropometric measures and osteoarthritis knee in non-obese subjects: a cross sectional study. Clinics 2011;66:275–9.doi:10.1590/S1807-59322011000200016
      OpenUrl
    47. ↵
      2. Stevens J ,
      3. McClain JE ,
      4. Truesdale KP
      . Selection of measures in epidemiologic studies of the consequences of obesity. Int J Obes 2008;32:S60–6.doi:10.1038/ijo.2008.88
      OpenUrlCrossRef
    48. ↵
      2. Greenland S ,
      3. Mansournia MA ,
      4. Altman DG
      . Sparse data bias: a problem hiding in plain sight. BMJ 2016;352:i1981.doi:10.1136/bmj.i1981
      OpenUrlFREE Full Text
    49. ↵
      2. Greenland S ,
      3. Mansournia MA
      . Penalization, bias reduction, and default priors in logistic and related categorical and survival regressions. Stat Med 2015;34:3133–43.doi:10.1002/sim.6537
      OpenUrlCrossRefPubMed

    Footnotes

    • Contributors DJC conceived and designed the study, distributed the survey, collected, analysed and interpreted the data, and drafted the manuscript. BES assisted with the conception of the study and design. BES critically revised the manuscript and gave final approval of the version to be published. MEB assisted with the conception of the study and with accessing the study participants. MEB critically revised the manuscript and gave final approval of the version to be published. DP assisted with the conception of the study and with the distribution of the survey. DP critically revised the manuscript and gave final approval of the version to be published. All authors read and approved the final manuscript.

    • Funding This work was supported by Arthritis Research UK (grant number 20194), funding gratefully received from Arthritis Research UK Centre for Sport, Exercise and Osteoarthritis.

    • Disclaimer The views expressed in the submitted article are those of the authors and not an official position of the institution or funder.

    • Competing interests None declared.

    • Patient consent Not required.

    • Ethics approval The Faculty of Medicine and Health Sciences Research Ethics Committee, The University of Nottingham (Ethics Reference No: K13022014). It was agreed in the ethics approval and with the British Olympic Association and the British Olympic Associations Athletes' Commission that implied consent would be used to avoid the use of personal signatures that could identify the GB Olympian. The letter of invitation and the participant information sheet made it explicit that by completing and returning the Olympian questionnaire, GB Olympians gave implied consent. GB Olympians were informed that if they completed and returned the Olympian questionnaire, they were consenting to have their details stored for future analysis. GB Olympians were informed that they were giving authorisation for inclusion of their data in public release data sets, such as journal articles and press releases.

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

    • Data sharing statement An anonymised summary of the data set generated and analysed during the current study may be available from the corresponding author on reasonable request.