Background The infrapatellar fat pad (IPFP) is of uncertain significance for knee osteoarthritis. The aim of this study was to describe the longitudinal associations between baseline IPFP maximal area and changes in knee pain, knee cartilage volume and cartilage defects in older adults.
Methods 356 community-dwelling male and female adults aged 50–80 years were measured at baseline and approximately 2.6 years later. T1-weighted or T2-weighted fat-suppressed MRI was used to assess maximal IPFP area, cartilage volume and cartilage defects at baseline and/or follow-up. Knee pain was assessed by the self-administered Western Ontario McMaster Osteoarthritis Index questionnaire.
Results After adjustment for confounders, IPFP maximal area in women was significantly and negatively associated with changes in knee pain (β: −0.18 to −0.86 for total knee pain, pain at night while in bed, pain when sitting/lying and pain when standing upright, all p<0.05) but not with other knee pain subscales. IPFP maximal area in women was beneficially associated with change in tibial cartilage volume per annum (β: +1.56% per cm2 at medial site; +0.86% per cm2 at lateral site, both p<0.05), but not with change in patellar cartilage volume. Further, it was significantly associated with reduced risks of increases in medial cartilage defects (relative risk: 0·46 at tibial site, relative risk: 0.59 at femoral site; both p<0.05) but not with increases at other sites in women. No significant associations were found in men.
Conclusions While the associations are not fully consistent, IPFP maximal area appears to have a protective role for knee symptoms and cartilage damage in older female adults.
- Knee Osteoarthritis
- Outcomes research
Statistics from Altmetric.com
Osteoarthritis (OA) is the most prevalent form of arthritis characterised by gradual cartilage loss, osteophyte formation and other joint structural changes,1 and is most common in the knees.2 Although the pathogenesis of OA is unclear, mechanical and metabolic factors have been shown to play roles in the initiation and progression of this disease.1 ,2 Age,3 sex4 and body mass index (BMI)5 are well-known risk factors for OA.
Infrapatellar fat pad (IPFP), an intracapsular but extrasynovial structure,6 is situated in the knee under the patella, between patellar tendon, femoral condyle and tibial plateau, and is structurally similar to subcutaneous adipose tissue.7 Recent studies8 have focused on the roles of immune cells, nerve fibres and adipocytes derived from IPFP, and considered IPFP as an active tissue in knee OA, but the results were not consistent9 ,10; moreover, IPFP is located close to the cartilage and bone surface so that it may reduce the impact of loading and absorb forces generated through the knee joint, and thus may have a beneficial role in knee OA. Indeed, our previous cross-sectional study has suggested that IPFP maximal area was not associated with total body fat, but had consistently beneficial associations with knee pain, cartilage defects and volume as well as radiographic OA (ROA)11; however, the cross-sectional results cannot delineate causal relationships between IPFP and OA.
Thus, the aim of this study was to describe the longitudinal associations between baseline IPFP maximal area and changes in knee pain, cartilage volume and cartilage defects in older adults.
Materials and methods
This study was carried out as part of the Tasmania Older Adult Cohort study. Subjects (98% Caucasian) between the ages of 50 years and 80 years were randomly selected using computer-generated random numbers from the roll of electors in southern Tasmania (population, 229 000), with an equal number of men and women, and the overall response rate was 57%. Baseline data were collected for 1100 participants, and follow-up data were approximately 2.6 years later (range 1.4–4.8 years). Follow-up MRI scans were only available for 404 participants due to decommissioning of the MRI machine. Institutionalised persons and subjects with contraindications to MRI were excluded. Self-report of disease status such as rheumatoid arthritis (RA) was recorded by questionnaire, and subjects with RA were excluded from analyses. Pain medication uses including non-steroid anti-inflammatory drugs, selective cyclooxygenase-2 (COX-2) inhibitors and analgesics in most days of the last month at baseline were recorded. The study was approved by the Southern Tasmanian Health and Medical Human Research Ethics Committee, and written informed consent was obtained from all participants.
Anthropometrics and joint pain assessment
Height was measured to the nearest 0.1 cm (with shoes, socks and headgear removed) using a stadiometer. Weight was measured to the nearest 0.1 kg (with shoes, socks and bulky clothing removed) by using a single pair of electronic scales (Delta Model 707, Seca, Hamburg, Germany) that were calibrated using a known weight at the beginning of each clinic.
The assessment of knee pain (walking on flat surface, going up/down stairs, at night while in bed, when sitting/lying and when standing upright) was self-reported, using the Western Ontario McMaster Osteoarthritis Index (WOMAC) with a 10-point scale from 0 (no pain) to 9 (most severe pain).12 Each component of joint pain was summed to create a total pain score (0–45), and prevalent knee pain was defined as a total score ≥1. Change in knee pain was calculated as (follow-up value − baseline value) for each subscale as well as total pain, and an increase in pain was defined as a change in score of 1 or greater, as we reported that the smallest statistically significant difference for the WOMAC knee pain score to be 0.9 for our population.13 ,14
Knee radiographic assessments
A baseline standing anteroposterior semiflexed view of the right knee with 15° of fixed knee flexion was performed and scored individually for osteophytes and joint space narrowing (JSN) on a scale of 0–3. The presence of ROA was defined as total score of ≥1 for JSN and osteophytes.14
Magnetic resonance imaging
MRI scans of the right knees were performed at baseline and follow-up. Knees were imaged in the sagittal plane on a 1.5-T whole body magnetic resonance unit (Picker, Cleveland, Ohio, USA) with the use of a commercial transmit-receive extremity coil. The following image sequences were used: (1) a T1-weighted fat saturation three dimensional (3D) gradient recall acquisition in the steady state; flip angle 30° ; repetition time 31 ms; echo time 6.71 ms; field of view 16 cm; 60 partitions; 512×512 matrix; acquisition time 11 min 56 s; one acquisition. Sagittal images were obtained at a partition thickness of 1.5 mm and an in-plane resolution of 0.31×0.31 (512×512 pixels). (2) a T2-weighted fat saturation 2D fast spin echo, flip angle 90°, repetition time 3067 ms, echo time 112 ms, field of view 16 cm, 15 partitions, 256×256-pixel matrix; sagittal images were obtained at a slice thickness of 4 mm with an interslice gap of 1.0 mm.
IPFP was measured by manually drawing disarticulation contours around the IPFP boundaries (figure 1) on a section-by-section T2-weighted MR image, using the software program Osiris (University of Geneva). The maximal area was selected to represent the IPFP size. One observer graded IPFP area in all MRIs. The intraclass correlation coefficient was 0.96 for intraobserver reliability (measured in 40 images by one observer), and interobserver reliability was 0.92 (measured in 40 images by two observers).
IPFP abnormalities including signal intensity alteration and soft tissue thickening were assessed (see online supplementary appendix 1).
Knee cartilage volume was determined on T1-weighted MR images with image processing on an independent workstation, as previously described.15 ,16 Total cartilage volume was divided into patellar, medial and lateral cartilage volumes by manually drawing disarticulation contours around the cartilage boundaries, section by section, which were then resampled for the final 3D rendering.17 The volume of the particular cartilage plate (patellar, medial tibial and lateral tibial) was then determined by summing all the pertinent voxels within the resultant binary volume. The intraclass correlation coefficients ranged from 0.92 to 0.96 for intraobserver reliabilities. Change in cartilage volume per annum in each site was calculated as: (follow-up cartilage volume − baseline cartilage volume)/(baseline cartilage volume×follow-up time).
Cartilage defects (0–4 scale) were assessed at the medial tibial, medial femoral, lateral tibial, lateral femoral and patellar sites using T1-weighted images as previously described16 and were further confirmed using T2-weighted images as follows: grade 0=normal cartilage; grade 1=focal blistering and intracartilaginous low-signal intensity area with an intact surface; grade 2=irregularities on the surface or bottom and loss of thickness <50%; grade 3=deep ulceration with loss of thickness >50%; grade 4=full-thickness chondral wear with exposure of subchondral bone. The presence of cartilage defect was defined as a cartilage defect score of ≥2 at one site. Intraobserver reliability was 0.89–0.94 and interobserver reliability was 0.85–0.93.16 An increase in cartilage defect score was defined as an increase of 1 or more on the 0–4 scale from baseline to follow-up at any site.
Tibial plateau bone area was determined by manually measuring on axial T1-weighted MR images, as previously described.17
Data were entered into a computerised data base using a blinded double-entry procedure, after which descriptive statistics for participants’ characteristics were tabulated. Scatter plots were used to examine the linear associations between baseline IPFP maximal area and changes in cartilage volume (including medial and lateral tibial) per annum. Univariable and multivariable log binomial regressions were used to estimate relative risk for the associations between IPFP area and increases in cartilage defects at five sites, and to estimate p values for increases in total knee pain, pain walking on flat surface, pain going up/down stairs, pain in bed at night, pain when sitting/lying, pain when standing, before and after adjustment for age, sex, height and weight. Univariable and multivariable linear regression analyses were used to examine the associations between IPFP area and change in cartilage volume or change in knee pain before and after adjustment for age, sex, height, weight and/or tibial bone area (only for cartilage volume). Standard diagnostic checks of model fit and residuals were routinely made, and data points with large residuals and/or high influence were investigated for data errors.
Interactions between sex and IPFP area on changes in outcome measures were investigated by regressing individual change in an outcome on a binary (0/1) term for sex within IPFP area, and assessed by testing the statistical significance of the coefficient of a (sex×IPFP area). We separated men and women for analysis because distinct sex differences were found in our results. A p value of <0.05 (two-tailed) or a 95% CI not including the null point (for linear regression) or 1 (for logistic regression) was considered statistically significant. All statistical analyses were performed on SPSS V.20·0 for Windows (SPSS, Chicago, Illinois, USA).
A total of 1100 subjects (51% women) aged between 51 years and 81 years (mean 62 years) participated in the Tasmania Older Adult Cohort study. Over 2.6 years, 271 subjects were lost to follow-up due to reasons as explained previously.18 In 829 subjects (85% of those originally studied) who completed the study, the first 404 subjects had a second MRI scan, including 40 subjects with RA who were excluded from analyses; so 356 participants were included in this longitudinal study. There were no significant differences in demographic factors and knee cartilage volume between subjects who had and who did not have the second MRI scans. The average time to follow-up was 2.6 years. Characteristics of participants are presented in table 1. Subjects with higher, middle and lower tertiles of IPFP size were similar in terms of age, knee pain and knee ROA; however, those with lower IPFP size had lower height, weight, cartilage volume and tibial bone area than those with larger IPFP size in men and women. Lower IPFP size was significantly associated with greater IPFP signal intensity alteration and soft tissue thickening (see online supplementary appendix figure S1).
Baseline IPFP maximal area and changes in knee pain
In women, IPFP maximal area was significantly and negatively associated with changes in total WOMAC knee pain and knee pain at night while in bed, when sitting/lying and when standing upright but not with knee pain when walking on flat surface and going up/down stairs after adjustment for potential confounders (table 2). It was also significantly associated with increases in total WOMAC knee pain and knee pain at night while in bed and when standing upright (figure 2A). In men, we did not find statistically significant associations between IPFP maximal area and changes or increases in total WOMAC knee pain and knee pain subscales (table 2, figure 2B).
Baseline IPFP maximal area and changes in knee cartilage volume
In multivariable analyses, IPFP maximal area was significantly and positively associated with changes in medial and lateral tibial cartilage volume per annum but not with change in patellar cartilage volume in women (table 3, figure 3A,B). IPFP maximal area was not significantly associated with changes in medial tibial, lateral tibial and patellar cartilage volume in men (table 3, figure 3C,D).
Baseline IPFP maximal area and increases in cartilage defects
In multivariable analyses, IPFP maximal area was significantly and negatively associated with increases in cartilage defects at medial tibial and femoral sites but not at lateral tibial and femoral and patellar sites in women (table 4). IPFP maximal area was not significantly associated with increases in cartilage defects at any sites in men (table 4).
The associations with changes in knee pain, changes in cartilage volume and increases in cartilage defects remained largely unchanged if the tertiles of IPFP maximal area or the normalised IPFP area (IPFP maximal area divided by total tibial bone area) was used as the independent variable (data not shown). They also remained largely unchanged after further adjustment for weight changes over 2.6 years and use of pain medications at baseline (data not shown).
To our knowledge, this is the first longitudinal study to examine the associations between IPFP size at baseline and changes in knee OA measures in community-based older adults. Our major findings were that higher IPFP maximal area at baseline was associated with reduced development and progression of knee pain, cartilage volume loss and cartilage defects over 2.6 years after adjustment for potential confounding factors. These associations were significant in women, not in men, suggesting IPFP size may have a beneficial role on knee symptoms and structure in women.
Recent research interests on IPFP are focusing on its potential metabolic and proinflammatory roles as an adipose tissue.7 IPFP is in close contact with synovial layers and articular cartilage, and can produce several cytokines locally into the knee joint.19 In patients with OA, IPFP secretes higher levels of inflammatory cytokines and adipokines including interleukin 6, adipsin, adiponectin and visfatin than subcutaneous fat.20 It also generates multiple oxylipins, and patients with OA can be distinguished from normal donors based on the secretion of lipid mediators involved in the oxylipin pathways by IPFP.21 These suggest that inflamed IPFP may play a detrimental role in the pathogenesis of OA.
Physiologically or at an early stage of OA, IPFP may have a different effect. It can contribute to the enlargement of the synovial area which could improve the distribution of lubricant in the knee joint.18 It may have buffering and lubricating functions in knee joints, which limits the knee's excessive activities, absorbs shocks from the anterior knee and reduces friction between the patellar tendon and the tibia, and thus is protective. IPFP is preserved even under extreme starvation conditions in which the subcutaneous adipose tissue is eliminated, indicating that this fat depot may be different from systematic fat and be critically important for knee function.22 We reported that IPFP area did not correlate with BMI and total body fat in older adults (Han et al, submitted) and another study reported that IPFP volume was not associated with BMI in healthy or OA subjects,23 indicating that systemic metabolic changes do not necessarily affect the size of IPFP or IPFP may not have metabolic and inflammatory effects physiologically at an early stage of OA.
The IPFP is regarded as an important source of pain in knee OA as it contains substance P fibres, which have been identified throughout the IPFP and surrounding synovial tissue7 ,24; however, IPFP resection has not been found to affect the function, range of motion and anterior knee pain in OA.25 Signal intensity alteration in the infrapatellar and intercondylar regions of IPFP, a surrogate for synovitis, was associated with fluctuations in knee pain in patients with knee OA,26 but the association between IPFP size and knee pain has not been elucidated. Our current study reported that the IPFP maximal area was significantly and negatively associated with changes in total WOMAC knee pain, knee pain at night while in bed, knee pain when lying/sitting and knee pain when standing upright. With every 1 cm2 increase in IPFP area, total knee pain score reduced by 0.86 over 2.6 years in women. This is clinically relevant, and suggests that IPFP area has a protective effect on mechanical and inflammatory knee pain, and reducing IPFP area in the process of knee surgery may increase later knee pain in women.
To gain insight into the role of IPFP in the joint, it is important to investigate its interaction with the other tissues known to be involved in the pathophysiology of OA. Previous in vitro studies report contradictory results.24 ,27 We reported that IPFP size was cross-sectionally associated with decreased knee JSN, increased knee cartilage volume and reduced prevalence of knee cartilage defects, in 977 older subjects.11 These indicated that IPFP size might have a protective role against cartilage loss but this needs to be substantiated by longitudinal studies. In the current study, we found that higher baseline IPFP area was associated with reduced loss of medial and lateral tibial cartilage volume, and lesser increases in medial tibial and femoral cartilage defects over 2.6 years in women. These were independent of potential confounders suggesting great IPFP size can reduce knee cartilage loss, particularly in the medial tibiofemoral compartment. With every 1 cm2 increase in IPFP area, medial tibial cartilage volume will increase by 1.6% per annum (or its loss will reduce by 1.6% per annum) in women. We did not find significant associations between baseline IPFP area and changes in patellar cartilage volume or defects over time, and the underlying reasons are unclear this time.
The reasons underlying the protective effects of IPFP maximal area on OA changes in women are unclear. IPFP maximal area was associated with decreased IPFP abnormalities including signal intensity alteration (that may imply inflammation and oedema) and soft tissue thickening (that may represent fibrosis), indicating that IPFP size-related decreases in pathological changes of IPFP may play roles. It may also be that some biochemical factors secreted from IPFP are protective, because IPFP-derived fat-conditioned medium in healthy individuals contained elevated levels of anti-inflammatory lipid mediator lipoxin A4,21 and IPFP fat-conditioned medium of patients with OA inhibited catabolic processes in cartilage.28 Additionally, it may be due to the presence of mesenchymal stem cells (MSCs) in IPFP, since MSCs from IPFP had a higher chondrogenic and adipogenic capacity than MSCs from normal cartilage and bone marrow,29 and inhibited secretion of proinflammatory cytokines/chemokines from OA synovium and chondrocytes.30 Another possible reason is that IPFP has a shock-absorbing role. Biomechanical factors, especially abnormal mechanical stress/loading, play an important role in the initiation and progression of OA.31 IPFP may have the same function as meniscus that can reduce the mechanical overloading and absorb the loading through the joint. Additionally, IPFP, having the same anatomical location as the patellar ligament, may reduce instability and the potential for injury to the joint, and thus prevent the onset and progression of OA. Based on these findings, we conjecture that IPFP, especially the size of it, may play a beneficial, rather than detrimental, role in the initiation and progression of OA. Currently, partial or complete excision of IPFP is commonly performed in knee surgery; however, the consequence of this procedure on clinical outcomes of OA is unknown.25 The results from our study suggest that this procedure may be worthy of reconsideration.
The strength of this study is that we studied a community-based population with a larger sample size using sensitive measures for 2.6 years. It has several potential limitations. This study measured IPFP area in 2D T2-weighted MRI, rather than IPFP volume in 3D T1-weighted MRI. However, the boundary of IPFP in 2D T2-weighted MRI is easier to define than that in 3D T1-weighted MRI; and IPFP area is highly correlated to IPFP volume in our hands (r=0.87). Measurement error may influence results. However, all measures were highly reproducible suggesting this is unlikely.
In summary, this longitudinal study suggests that IPFP maximal area appears to have a protective role for knee symptoms and cartilage damage in older female adults. IPFP with normal quality should be preserved during knee surgery.
The authors thank the participants who made this study possible, and gratefully acknowledge the role of the staff and volunteers in collecting the data, particularly research nurses Boon C and Boon P. Warren R assessed MRIs, and Dr Srikanth V and Dr Cooley H assessed radiographs.
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
FP and WH contributed equally.
Contributors CD had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design: CD, GJ and FC; Acquisition of data: FP, WH, XW, ZL, XJ, BA and CD; Analysis and interpretation of data: FP, WH and CD; Manuscript preparation: FP, WH, XW, ZL, XJ, BA, FC, GJ and CD; Statistical analysis: FP, WH and CD.
Funding This study was supported by the National Health and Medical Research Council of Australia (302204); Arthritis Foundation of Australia (MRI06161); Tasmanian Community Fund (D0015018); Masonic Centenary Medical Research Foundation; Royal Hobart Hospital Research Foundation; and University of Tasmania Institutional Research Grants Scheme (D0015019).
Competing interests GJ is supported by a National Health and Medical Research Council Practitioner Fellowship. CD is supported by an Australian Research Council Future Fellowship.
Patient consent Obtained.
Ethics approval The Southern Tasmanian Health and Medical Human Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.
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.