Objective To investigate whether offspring having at least one parent with a total knee replacement for severe primary knee osteoarthritis (OA) have an increased risk of worsening knee pain over 8 years as compared with controls with no family history of knee OA.
Methods A total of 219 participants (mean age 48 years, range 29–61 years) with 115 offspring and 104 controls participated in this study. Knee pain was respectively assessed using Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) at 2 years and 10 years. T1-weighted or T2-weighted fat saturated MRI of the right knee was performed to assess knee cartilage defects, bone marrow lesions, effusion, meniscal extrusion and tears.
Results Compared with controls, the prevalence of knee pain for offspring was similar at 2 years (56% vs 54%, p=0.764) and higher at 10 years (74% vs 54%, p=0.002). Over 8 years, offspring more frequently had an increase in total knee pain (66% vs 41% ≥1 point increase, p=0.003) and in all subscales apart from walking (all p<0.05). In multivariable analysis, after adjustment for confounders and structural factors, offspring had an elevated risk of worsening total knee pain (OR=2.16, 95% CI 1.14 to 4.12), as well as each subscale except for walking and standing (OR=1.95 to 3.30, all p<0.05).
Conclusions Offspring with a family history of knee OA have an increased risk of worsening knee pain, which is independent of structural factors, suggesting that genetic factors may be involved in the pathogenesis of knee pain.
- Knee Osteoarthritis
- Magnetic Resonance Imaging
Statistics from Altmetric.com
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.
Osteoarthritis (OA) is the most common form of arthritis in western countries.1 The knee joint is the major site of OA with a prevalence of 30% in those aged 65 years and above.2 Pain in the knee is the most common presenting symptom of knee OA, but it also occurs due to other musculoskeletal diseases.3 ,4 Knee pain leads to significant restrictions in function that prevent patients with OA from engaging in their daily activities,5 and may eventually require surgical treatment.6 It is reported that approximately 21–35% of people aged 45 years or over have had persistent knee pain lasting for at least 1 week during a month period.3 ,4 ,7 Although the mechanism of knee pain is not fully clear, many factors have been shown to be associated with knee pain including demographic, structural, genetic and central factors (such as pain-coping strategies or beliefs about knee pain).8–14
Factors outside knee joint structures appear to play a role in pain. Previous twin and sibling pair studies have demonstrated that genetic factors are involved in the pathogenesis of pain, with heritability estimates of approximately 50% for different pain traits.15 ,16 In addition, recent candidate gene studies have investigated gene polymorphisms associated with pain sensitivity,17–20 some of which can discriminate those with painful OA from those without pain. In a previous cross-sectional report we found a higher prevalence of knee pain in people with at least one parent undergoing a total knee replacement (TKR) for severe primary knee OA at baseline compared with controls, which was independent of structural factors, suggesting a possible role of genetic factors in knee pain.2 So far, there are no studies examining if people with a family history of knee OA have an increased risk of knee pain over time. The aims of this study were, therefore, to describe whether offspring of people who had at least one parent with TKR for severe primary knee OA would have an increased risk of worsening knee pain over 8 years as compared with controls with no family history of knee OA.
Materials and methods
This study was carried out in southern Tasmania in the capital city of Hobart. The initial measurements were taken from June 2000 to December 2001, a total of 372 participants (186 offspring and 186 controls) aged 26–61 years (mean age of 45 years) were enrolled. Phase 2 was conducted 2 years (range 1.8–2.6 years) later, 326 participants (162 offspring and 164 controls) were traced. Phase 3 was conducted at 10 years (range 9.1–11.4 years), 219 participants (115 offspring and 104 controls) aged 36–71 years were included (figure 1). Participants were selected from two sources, as described previously.2 ,21 ,22 Half of the participants were the adult children (offspring) of participants who had had a TKR performed for primary knee OA at any Hobart hospital from 1996 to 2000. This diagnosis was confirmed by reference to the medical records of the orthopaedic surgeon and the original radiograph where possible. The other half were controls selected at random from the state electoral roll (2000), without a history of knee OA in either parent. Participants from either group were excluded on the basis of contraindication to MRI (including metal sutures, presence of shrapnel, iron filing in eye and claustrophobia) and common rheumatoid diseases (rheumatoid arthritis and inflammatory arthritis). All participants provided informed written consent.
Knee pain was assessed by a self-reported pain questionnaire (yes or no pain) at baseline and was defined as pain for more than 24 h in the last 12 months, or daily pain on more than 30 days in the last year. The Western Ontario McMaster Osteoarthritis Index (WOMAC) was used to assess knee pain at 2 years and 10 years. The subscales of WOMAC which consist of five items (walking on flat surface, going up/down stairs, at night in the bed, sitting/lying and standing upright) on a 10-point scale ranging from 0 (no pain) to 9 (most severe pain) were used for this study.23 Each item was summed to produce a total pain (0–45) score with higher scores indicating greater pain. A total score of 1 or greater was considered as presence of knee pain at 2 years and 10 years.
Weight was measured to the nearest 0.1 kg (with shoes, socks and bulky clothing removed) using a single pair of electronic scales (Seca Delta Model 707) calibrated using a known weight at the beginning of each clinic. Height was measured to the nearest 0.1 cm (with shoes and socks removed) using a stadiometer. Body mass index (BMI) (kg/m2) was calculated.
A standing anteroposterior semiflexed view of the right knee was performed at baseline in all participants and scored individually using the Altman atlas for osteophytes and joint space narrowing (JSN) as previously described.24 The presence of radiographic OA (ROA) was defined as any score of 1 or greater for JSN or osteophytes.
Magnetic resonance imaging
A MRI of the right knee was performed with a 1.5 T whole-body magnetic resonance unit (Picker, Cleveland, Ohio, USA) using a commercial transmit-receive extremity coil at 2 years. As previously described,24 ,25 a T1-weighted fat suppression three-dimensional gradient recall acquisition and T2-weighted fat saturation two-dimensional fast spin echo acquisition were used. Cartilage defects were graded for medial tibial, medial femoral, lateral tibial, lateral femoral and patellar sites using a 0–4 point scale, as previously described,21 and the scores at these five sites were summed to create a total score of cartilage defect. The presence of any cartilage defect was defined as a score of greater than 1 at any site. Effusion was assessed in the suprapatellar pouch on T2-weighted MRIs using a 0–3 point scale,26 the presence of effusion was defined as a score of greater than 1. Bone marrow lesions (BMLs) were assessed on T2-weighted MRIs and defined as areas of increased signal adjacent to the subcortical bone, as previously described.25 The maximum area (cm2) of the lesion of different sites was measured, and the BML with the largest size was recorded if more than one lesion was present at the same site. The presence of BML was defined as a score of greater than 0 at any site. As previously described,27 ,28 meniscal tears and extrusion were evaluated within six defined regions (anterior horn, body, and posterior horn of each of the medial and lateral tibiofemoral compartments) using a 0–2 point scale. Meniscal tears and extrusion are often correlated and to ensure the least loss of participants,27 ,29 ,30 any meniscal pathology was created to be a dichotomous variable with a combination of meniscal tears and extrusion together in this study. The presence of meniscal pathology was defined as a score of 1 or greater of meniscal tears or extrusion at any site.
Knee pain was assessed at baseline by a simple questionnaire, not by WOMAC questionnaire, so a change in WOMAC knee pain as the main outcome was only available from phase 2 to phase 3. Change in WOMAC was calculated as (phase 3 value − phase 2 value) for total pain as well as each subscale. The smallest detectable difference for the WOMAC knee pain score was calculated to be 0.8 for our population,31 so an increase in score of 1 or greater was defined as the cut-off for worsening knee pain. Subscale-specific knee pain was defined as knee pain within the same subscale (eg, knee pain while walking on a flat surface and change in knee pain while walking on a flat surface). t Tests and χ2 tests were used to compare differences in means and percentages where appropriate. Mann–Whitney U test was used to compare absolute change in knee pain between the offspring and the control group. Logistic regression modelling was used to assess the potential relationships between the status of participants (offspring or controls) and change in knee pain (increase vs no increase) over 8 years, after adjustment for age, sex, BMI, smoking history and knee pain at baseline. Further adjustment for structural factors of relevance to pain in this cohort was also performed. Inverse probability weighting was used to examine whether loss to follow-up biased our results. p Values less than 0.05 (two-tailed) or 95% CIs not including the ‘1’ point were regarded as statistically significant. All statistical analyses were performed using SPSS Statistics (V.20; Chicago, Illinois, USA).
At 10 years of follow-up, a total of 219 participants comprising 115 offspring and 104 controls completed the study. Figure 1 describes the study population. Two hundred and seven (107 offspring and 100 controls) participants had complete WOMAC pain score information at phase 2 and phase 3. There were no statistically significant differences in characteristics of participants between the participants included in the current study and those lost to follow-up except for history of smoking (42% vs 59%).
The characteristics of included participants are presented in table 1. Offspring were heavier than controls, and had a higher percentage of smokers. No statistically significant differences were observed between the two groups in cartilage defects, meniscal pathology, effusion, BMLs and ROA. At baseline, offspring had a higher prevalence of knee pain, no statistically significant difference at phase 2; and a significantly higher prevalence at phase 3 compared with controls. In unadjusted analyses, offspring had higher knee pain scores while walking and climbing stairs at phase 2, and had a higher total knee pain score which approached but did not reach statistical significance. Pain scores in total knee pain and each subscale were consistently greater in magnitude than those of the controls at phase 3.
Changes in knee pain over 8 years between offspring and controls are presented in figure 2. There were increases over time in both groups, but these were greater in the offspring for all categories (changes in magnitude and increases) apart from knee pain on a flat surface and climbing stairs. No statistically significant difference was found in change in total WOMAC and other subscales of WOMAC (stiffness and physical function) (data not shown).
Table 2 describes the associations between offspring-control status and risk of any increase in knee pain. In univariable analyses, offspring status was associated with an increase in total knee pain and all subscales with the exception of knee pain on a flat surface. In multivariable analyses, the associations between the status of participants and increases in knee pain remained statistically significant after adjustment for age, sex, BMI, smoking history and baseline knee pain. After further adjustment for cartilage defects, BMLs, meniscal pathology and effusion, these associations persisted for total knee pain and all subscales apart from knee pain on a flat surface and standing, of which the latter was of borderline significance. Intriguingly, a trend to a dose-response relationship was found between the number of parents with TKR and change in total knee pain and each subscale. Numerically greater ORs were seen in those people with two parents undergoing TKR than those offspring with one parent with TKR (see online supplementary table S1). Consistent results were observed after re-analyses of data using inverse probability weighting.
This study found that offspring of those with severe knee OA had an increased risk of prevalent pain and worsening knee pain over 8 years as compared with controls who had no family history of OA, and this relationship persisted after adjustment for potential confounding factors and for joint structural abnormalities of relevance to pain. This implies that the genetic contribution to knee pain may be mediated through factors outside the joint possibly involving pain processing.
To date, there have been limited studies investigating the role of family history of knee OA on knee pain. Previously we reported, in a larger sample, that offspring had a higher prevalence of knee pain as assessed by a simple questionnaire at baseline.2 Consistent with this, the current study also found statistically significant differences in knee pain scores (prevalence and severity) assessed by WOMAC at 10 years. Surprisingly, there was no difference in the overall prevalence of knee pain at 2 years although there were higher pain scores while walking on a flat surface and climbing stairs. This variation may reflect the use of different questionnaires or more likely imply a greater effect on incident pain.
Pain experience in knee OA is a complex feature and the underlying aetiology of knee pain is multifactorial.14 Many earlier studies have shown that older age,3 ,11 ,12 female sex,3 ,32 ,33 previous knee injury34 ,35 and smoking35 ,36 appear to be important risk factors for developing knee pain. Also, overweight or obesity shows a causal relationship with the development of knee pain and knee OA.35 ,37–39 Despite structure-symptom discordance in radiographs, knee structural abnormalities in MRIs have been consistently associated with knee pain.14 ,40 The present study found greater weight, more former smokers and less knee injury in the offspring group; however, in this study, after adjustment for these factors and structural factors, the association with knee pain remained largely unchanged, suggesting that the differences cannot be explained by these factors.
Genetic predisposition to the development of knee pain appears important. The findings of the present study that offspring have elevated risk of worsening knee pain, with around a twofold higher risk than controls for total knee pain as well as subscales suggest underlying genetic components in worsening knee pain. These results are consistent with previous twin studies41 and earlier reports from our group which found the heritability of knee pain was higher in sib pairs.16 However, whether family factors such as pain-coping strategies, traditions and beliefs about knee pain have a role is unclear.12 ,42
OA-related pain is a complex integration of sensory, affective and cognitive processes,43 driven by nociceptive and neurobiological mechanisms,44 each of which involves a number of proteins throughout the peripheral and central nervous systems, whose effects have been shown to be affected by the interplay between environmental and genetic factors.41 ,45 Several studies have documented that genetic mutations can confer hypersensitivity or insensitivity to pain stimuli.46 ,47 Therefore, it is plausible that genetic factors may have a role in the pain sensitivity. Recent studies have attempted to examine the relationship between genes associated with pain sensitivity and OA-related pain.17–19 One study identified a genetic variant (Val158Met) in the catechol-O-methyltransferase gene involved in pain sensitivity to be associated with hip pain among those with hip OA17 but this was not confirmed in independent cohorts.20 Subsequently, Valdes et al18 found allelic variation in the lle585Val variant for the gene encoding transient receptor potential cation channel, subfamily V, member 1 was able to discriminate those with and without painful OA. In another study, the single nucleotide polymorphism in the proprotein convertase subtilisin/kexin type 6 gene also showed a strong protection against pain in those with knee OA.19 When combined with the lack of clear replication for radiographic OA at this point in time, these studies imply that specific genes and/or other multiple extra-articular factors may be more important in the pathogenesis of knee pain than radiographic OA,14 ,48 and that higher risk of worsening knee pain for offspring might be attributed to a difference in pain processing in the offspring.
The current study has several potential limitations. First, the proportion followed up was 59% at 10 years, so participants lost to follow-up may lead to bias; however, re-analyses of data using inverse probability weighting did not change any of the results, indicating robust results. Second, knee pain was measured using different methods—simple questionnaire at baseline and WOMAC at 2 years and 10 years—so we are unable to directly compare baseline with later phases. Both methods to assess pain may result in recall bias due to variation in reporting of pain,49 especially for offspring with family history of knee OA. Third, knee pain may result from other musculoskeletal diseases or other sites of the body;3 ,50 however, we have not screened for these conditions and did not evaluate pain in other sites. Lastly, several variables such as ROA were not measured at 2 years because of the perceived insensitivity of X-ray to detect radiological changes over this short period; it is unlikely that radiological changes would be different during that period.
In summary, this longitudinal study identified that offspring of people with a TKR for severe primary knee OA have an increased risk of worsening knee pain compared with controls and this relationship is independent of knee structural factors, suggesting that genetic factors may be involved in the pathogenesis of knee pain in middle life.
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:
- Data supplement 1 - Online supplement
Handling editor Tore K Kvien
Contributors FP participated in the design of the study, analysis and interpretation of the data and manuscript preparation. CD participated in interpretation of the data and revising the manuscript. TW participated in interpretation of the data and revising the manuscript. HK participated in acquisition of data and revising the manuscript. JM-P participated in acquisition of data and revising the manuscript. J-PP participated in acquisition of data and revising the manuscript. FC participated in the design of the study and revising the manuscript. GJ participated in the design of the study, interpretation of the data and manuscript preparation. All authors read and approved the final manuscript.
Competing interests None.
Patient consent Obtained.
Ethics approval This study was approved by the Southern Tasmanian Health and Medical Human Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.