|Home | About | Journals | Submit | Contact Us | Français|
In osteoarthritis (OA) the synovium is often inflamed and inflammatory cytokines contribute to cartilage damage. Omega-3 polyunsaturated fatty acids (n-3 PUFAs) have anti-inflammatory effects whereas omega-6 polyunsaturated fatty acids (n-6 PUFAs) have, on balance, proinflammatory effects. The goal of our study was to assess the association of fasting plasma phospholipid n-6 and n-3 PUFAs with synovitis as measured by synovial thickening on contrast enhanced (CE) knee MRI and cartilage damage among subjects in the Multicenter Osteoarthritis Study (MOST).
MOST is a cohort study of individuals who have or are at high risk of knee OA. An unselected subset of participants who volunteered obtained CE 1.5T MRI of one knee. Synovitis was scored in 6 compartments and a summary score was created. This subset also had fasting plasma, analyzed by gas chromatography for phospholipid fatty acid content, and non-contrast enhanced MRI, read for cartilage morphology according to the WORMS method. The association between synovitis and cartilage morphology and plasma PUFAs was assessed using logistic regression after controlling for the effects of age, sex, and BMI.
472 out of 535 subjects with CE MRI had complete data on synovitis, cartilage morphology and plasma phospholipids. Mean age was 60 years, mean BMI 30, and 50% were women. We found an inverse relation between total n-3 PUFAs and the specific n-3, docosohexanoic acid with patellofemoral cartilage loss, but not tibiofemoral cartilage loss or synovitis. A positive association was observed between the n-6 PUFA, arachidonic acid, and synovitis.
In conclusion, systemic levels of n3 and n6 PUFAs which are influenced by diet, may be related to selected structural findings in knees with or at risk of OA. Future studies manipulating the systemic levels of these fatty acids may be warranted to determine the effects on structural damage in knee OA.
Osteoarthritis (OA) is highly common in older adults and few treatments are available none of which target structural abnormalities in the joint (1, 2). Aside from cartilage and bone, OA affects ligaments, muscles, and synovium. OA is thought to be an inflammatory disorder with low grade inflammation affecting the synovium and inflammatory cytokines contributing to the cartilage damage which is the signature pathologic feature of disease (3). Ayral et al reported that synovitis tends to cause cartilage loss in the region where the synovitis occurs (4).
In previous work, we found a high prevalence of knee synovitis as measured by synovial thickening in contrast enhanced (CE) knee MRI in the Multicenter Osteoarthritis Study (MOST) (5). Synovitis of at least a mild degree was found in 56% of those with no radiographic knee OA and 67% of those with radiographic knee OA.
Omega-6 and omega-3 polyunsaturated fatty acids (n-6 and n-3 PUFAs) are directly linked to inflammation via their role as precursors for a family of compounds known as eicosanoids. Eicosanoids are mediators and regulators of inflammation. Arachidonic acid (AA) is the primary n-6 PUFA found in inflammatory cells and leads to the production of inflammatory eicosanoids. Studies have shown increasing the amount of either AA or its precursor linoleic acid (LA;18:2n-6) in the diet increases the content of AA in inflammatory cells and the production of inflammatory eicosanoids (6). Higher levels of AA in the blood and or dietary intake have been tied to more platelet aggregation and increased atherosclerosis in a subpopulation with a genetic variation in 5-lipoxygenase, an enzyme that generates inflammatory leukotrienes from AA (7-9). In rheumatoid arthritis, a diet low in AA ameliorates clinical signs of inflammation (10).
The n-3 PUFA, eicosapentaenooic acid (EPA;20:5n-3) and docosahexaenoic acid (DHA;22:6n-3) give rise to eicosanoids with a slightly different structure from those formed from AA. The functional significance is that these eicosanoids are less potent mediators of inflammation. Increased consumption of EPA and DHA results in increased proportions of these fatty acids in inflammatory cells and occurs in a dose response fashion and at the expense of AA (11, 12). So in addition to forming less inflammatory eicosanoids, less substrate is available for the formation of AA derived eicosanoids.
In multiple trials treatment with n-3 PUFAs, predominantly EPA and DHA, resulted in lower levels of proinflammatory markers and higher levels of antiinflammatory markers (13, 14) and better outcomes in numerous diseases including RA (15, 16) and cardiovascular disease (17-19). Also, in cartilage cell cultures n-3 PUFAS have been shown to inhibit the transcription of major enzymes and cytokines tied to matrix degradation (20).
The precursors to AA and (EPA and DHA), LA;18:2n-6 and alpha linolenic acid (ALA; 18:3n-3) respectively, are essential in the diet. The predominant dietary n-6 PUFA is LA that is converted to AA, in addition some AA is obtained by meat consumption. The predominant dietary n-3 PUFA is ALA, and research shows that only a very small portion is converted to the long chain n-3 EPA and DHA that seem to be the more potent inhibitors of inflammation (21). Sources of EPA and DHA in the diet include fatty fish like salmon, tuna, anchovies, sardines and shellfish, shrimp and Alaskan king crab. The current Western diet is higher in n-6 fatty acids with a n-6 to n-3 ratio of 20-30 to 1 (22). This ratio has increased dramatically over the years with food processing, less fish consumption, and the dietary habits of our farm animals.
The goal of our study was to assess the association of fasting plasma phospholipid n-6 (AA) and n-3 PUFAs (EPA and DHA) with synovitis as measured by synovial thickening on contrast enhanced (CE) knee MRI among subjects in MOST. Since cartilage loss on MRI may also reflect the influence of local inflammation, which may, in turn, be related to systemic inflammation, we also examined the relation of cartilage loss with these n-6 and n-3 PUFA levels. Potential findings could lead to dietary recommendations for fatty acid intake to ameliorate inflammation related to osteoarthritis.
The Multicenter Osteoarthritis Study (MOST) is the parent study for this investigation of knee pain and synovitis and has been described elsewhere (5). All subjects were recruited from two U.S. communities, Birmingham, Alabama and Iowa City, Iowa. Subjects were excluded if they screened positive for RA, had ankylosing spondylitis, psoriatic arthritis, chronic reactive arthritis, had renal problems resulting in a need for hemo- or peritoneal dialysis, a history of cancer (except for nonmelanoma skin cancer), bilateral knee replacement surgery, inability to walk without the help of another person or walker, or were planning to move out of the area in the next three years. Persons who had no contraindications to MRI and who were willing obtained extremity-based MRI’s (see below) repeatedly throughout the study. The protocol for MOST was approved by IRB’s at University of Iowa, University of Alabama, Birmingham, University of California, San Francisco and Boston University Medical Center.
For this investigation on synovitis, a subset of MOST subjects with the following inclusion criteria were investigated: volunteered to undergo a 1.5T contrast enhanced MRI of one knee at the 30 month follow-up clinic visit. Radiographs had already been obtained and read and we selected the knee with the lower Kellgren Lawrence grade for contrast enhanced MRI. If the grade was the same for both knees, the dominant leg was chosen. Subjects with Kellgren and Lawrence grade of 4 in both knees were ineligible. For those subjects with kidney disease, diabetes or over the age of 65, a serum creatinine was obtained and the glomerular filtration rate calculated before gadolinium injections. Those subjects with renal insufficiency were excluded from this substudy.
Contrast enhanced MRIs were obtained at 30 months with a 1.5-T system (Siemens Symphony) with a circumferential extremity coil. Axial and sagittal T1-weighted contrast enhanced sequences were acquired (TR=600ms, TE=13 ms, 3.0 slice thickness, 0.3 mm interslice gap, FOV 16 × 16 cm, matrix 512 × 512, ETL 1). Intravenous gadolinium (Magnevist® (gadopentetate dimeglumine) or Omniscan® (gadodiamide)) was administered at a dose of 0.2ml (0.1mmol)/kg body weight. Two minutes after completing the injection of the gadolinium, sagittal sequences were obtained immediately followed by the axial sequences.
The MOST parent study MRIs were also performed at 30 months on a 1.0 T extremity based Orthone magnet (OrthOne™, Oni Inc., Wilmington, MA) with a circumferential extremity coil for which contrast enhanced scans were not advisable. These MRI’s were acquired using fat suppressed fast spin echo proton density weighted sequence in two planes, sagittal (TR 4800 TE 35 slice thickness 3mm, 0mm interslice gap, FOV 14cm, 288 × 192 matrix, NEX 2), and axial (TR 4700 TE 13.2 slice thickness 3 mm, 0 mm interslice gap, FOV 14 cm, and 288 × 192 matrix, NEX 2) and a STIR sequence in the coronal plane (TR 7820 TE 14 TI 100, slice thickness 3mm, 0mm interslice gap, 14 cm, 256 × 256 matrix , NEX 2).
For the MRI readings we equated synovial thickening with synovitis. Using a protocol described elsewhere (5), synovitis on contrast enhanced MRI using axial and sagittal T1-weighted sequences was scored separately at 6 sites (medial and lateral parapatellar recesses, suprapatellar pouch, infrapatellar fat pad, and medial and lateral posterior condyles). Degree of synovitis was scored semiquantitatively (0-3) in the medial and lateral parapatellar recesses, suprapatellar pouch and the infrapatellar fat pad. For the posterior medial and lateral femoral condyles, we scored synovitis as either present (1) or absent (0). This semiquantitative scoring method has been validated in contrast enhanced MRI (23). One reader (KB), trained by a musculoskeletal radiologist (AGr) and blinded to pain scores, read synovitis on contrast enhanced MRIs.
As previously reported (5) to obtain a summary score for the 6 regions where synovitis was scored on contrast enhanced MRI, we created the following categories; 1) normal or questionable; ≤ 4 sites scored as 1 and all other sites scored as 0, 2) mild; ≥ 4 sites scored as 1 and/or ≤1 site scored as 2, 3) moderate; ≥ 2 sites scored as 2 and no score of 3, 4) severe; ≥ 1 site scored as 3. Due to small numbers of MRI’s classified as ‘moderate’ and ‘severe’ synovitis, these categories were collapsed in the analysis.
To obtain information about other MRI features, we used the non-contrast enhanced MRI. Two musculoskeletal radiologists (AG and FR) read non-contrast enhanced MRIs for cartilage morphology according to the Whole-Organ Magnetic Resonance Imaging Score (WORMS) method (24). In each of 10 subregions within the medial and lateral tibiofemoral compartments and in 4 subregions of the PF compartment, cartilage morphology was scored from 0 to 6 where 0 is normal morphology and 6 is diffuse cartilage loss to bone. To classify the amount of cartilage loss, we created 3 categories: normal morphology (grades 0 and 1), erosions without more diffuse cartilage loss (grades 2 and 3) and diffuse loss (>=4). To define the amount of cartilage loss in a region (e.g. tibiofemoral compartment), we took the maximal score of any subregion (e.g. central medial femur) within that region.
Lipids were extracted from plasma (25) drawn at the study baseline after addition of an internal standard (25 μg of 1,2 diheptadecanoyl-glycero-3-phosphocholine). The phospholipid subfraction was separated by solid-phase extraction using aminopropyl columns (26), saponified and then methylated (27). The supernatant containing the fatty acid methyl esters (FAME’s) was dried down under nitrogen, resuspended in 100ul of hexane, transferred into amber GC vials and stored at −20 °C until the time of analysis.
The phospholipid FAMEs were analyzed by an Autosystem XL gas chromatograph (Perkin Elmer, Boston MA) equipped with a 30m x 0.25mm i.d (film thickness 0.25μm) capillary column (HP INNOWAX, Agilent Technologies, DE) as previously described (18, 28). Peaks of interest were identified by comparison with authentic fatty acid standards (Nu-Chek Prep, Inc. MN) and expressed as molar percentage (mol %) proportions of fatty acids relative to the internal standard. Pooled plasma samples used as additional controls were run weekly. On average the coefficient of variations range from 0.5 to 4.3% for fatty acids present at levels > mol%, 1.8 to 7.1% for fatty acids present at levels between 1-5 mol%, and 2.8 to 11.1% for fatty acids present at levels <1 mol%.
We evaluated the cross-sectional association between synovitis and cartilage morphology and plasma phospholipid fatty acids (AA, EPA, DHA, and total n-6 and n-3 PUFAs) using logistic regression. We controlled for the effects of age, sex, BMI on synovitis. In these analyses, the MRI feature was treated as a dichotomous dependent variable and fatty acid category was the independent variable. When there were three levels of the MRI feature (e.g. synovitis), we carried out two separate logistic regressions, each using normal/questionable as one of the dichotomous categories. We also performed proportional logistic regression (which uses all the data in one analysis but did not allow us to use the nonsynovitis or normal cartilage group as the sole referent as a dependent variable) and came up with similar findings. All analyses were performed using SAS 9.1 (SAS Institute, Cary, NC). Due to possible confounding by indication we initially excluded those taking fatty acid supplements (n=17). However, results were the same when these 17 persons were left in the analysis. Results presented include the 17 taking supplements.
The sample studied consisted of equal numbers of men and women (see table 1) with a mean age of 60 years. Approximately one-third showed evidence of x-ray OA in the studied knee and about 2/3’s had signs of synovitis on MRI.
We examined total n-6 PUFAs and found no association with either synovitis or cartilage loss (data not shown). But when we examined AA levels, we found that higher levels of this n-6 PUFA were associated with more synovitis (table 2a) although not with cartilage loss (table 2b).
We next examined total n-3 PUFAs and their relation to structural findings in the knee and found that there was no association with synovitis (table 3a), but higher levels of total n-3 PUFAs were associated with less severity of patellofemoral (table 3b), but not tibiofemoral cartilage loss. Similarly, for DHA we found no association with synovitis (table 4a) but did find a relation of higher DHA levels with less patellofemoral cartilage loss (table 4b) although not with tibiofemoral loss. For EPA levels, another n-3, we found no association with either synovitis or cartilage loss in any compartment (data not shown). Results were similar whether the PUFAS were analyzed in quartiles or tertiles.
Our results, examining fasting plasma levels of n-6 and n-3 PUFAs and OA, suggest subjects in MOST with high levels of total n-3 PUFAs (in the fourth quartile), specifically DHA, had lower patellofemoral cartilage loss. In contrast, MOST subjects with higher levels of AA tended to have higher synovitis severity.
We are unaware of other studies that have looked in vivo at the relation of n-6 and n-3 PUFA levels with knee structures in OA. There have been studies on the association of fatty acids with OA in general or elements related to disease. For example, in a study looking at fat in OA bone, levels of n-6 PUFAs, especially AA was almost double that seen in a comparator osteoporotic bone (29). Also, in patients undergoing joint replacement surgery for OA, those taking fish oils (rich in n-3 PUFAs) before surgery had a near doubling of long chain fatty acids in synovial fluid and bone marrow compared to before supplementation (30). AA has been shown to be increased in cartilage, serum, and synovial fluid in OA patients compared to non OA (31).
We found an effect of both total n-3 PUFAs and of DHA on patellofemoral cartilage loss. These two analytes are not synonymous, with DHA representing only a small fraction of total n-3 PUFAS but representing one of the biologically active fractions. Our results are not unique in that others examining n-3 PUFA relations with disease have also reported an association uniquely with DHA (32).
Why would there be an effect of relevant fatty acids on patellofemoral but not tibiofemoral cartilage? We are not sure but note that Ayral et al (4) in a longitudinal arthrosopy study reported that cartilage loss occurred only adjacent to active synovitis and not distant from it. Synovitis predominates in the parapatellar area (33). In our study subset, all subjects with moderate to severe synovitis had this inflammation in the parapatellar region. Therefore, it is possible that the patellar cartilage loss we saw was due to this adjacent synovits and that synovitis did not invest the tibiofemoral compartment.
We acknowledge that our findings are not consistent across n-6 and n-3 PUFAs or their components and some of our results are null. The inconsistent and null findings may partially be explained by the observational nature of the study. The levels of n-3 PUFAs did not attain levels reached in clinical trials where subjects are supplemented with n-3 PUFAs. As mentioned earlier, the Western diet is much higher in n-6 than n-3 PUFAs thereby reducing the potential antiflammatory benefits of n-3 PUFAs. In some trials the intake of n-6 PUFAS was restricted in conjunction with n-3 supplementation, and in these trials an enhanced benefit for n-3 PUFAs was seen (34, 35).
We interpret our findings as indicating a potential positive effect on synovitits and cartilage loss of n-3 PUFAs and negative effect of n-6 PUFAs, specifically AA. As the first descriptive observational study of these fatty acids and their relation to relevant structural abnormalities, we hope that these results generate further investigations that may clarify our findings.
There are several limitations when interpreting our data. We were only able to obtain one measurement of fatty acid intake at baseline. The measurement will be affected by current fatty acid intake and there is the potential for misclassification. We cannot make causal inferences because we have not manipulated the diet and we do not have longitudinal data. And, an observational association of health behaviors, such as dietary fatty acid intake, with disease may only indicate an indirect connection (ie. those with a healthy diet also exhibit other behaviors that are beneficial to their joints).
In conclusion, this observational study suggests that future studies manipulating the systemic levels of n-6 and n-3 PUFAs, which are influenced by diet, may be warranted to determine the effects of these fatty acids on synovitis and cartilage loss in knees with or at risk of OA.
Supported by MOST U01’s AG18947; AG18832; AG19069; AG18820, and by NIH AR47785 and K23AR053855
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Author contributions: Kristin Baker, design of the study, acquisition of the data, analysis and interpretation and drafting of the article. Nirupa Matthan, analysis and interpretation of the data and revising the article. Alice Lichtenstein, interpretation of the data and revising the article. Jingbo Niu, analysis and interpretation of the data and revising the article. Ali Guermazi and Frank Roemer, acquisition of the data and revising the article. Andrew Grainger, acquisition of the data, analysis and interpretation of the data, and revising the article, Michael Nevitt, acquisition of the data, analysis and interpretation of the data and revising the article. Margaret Clancy, acquisition of the data and revising the article. Cora Lewis and James Torner, acquisition of the data and revising the article. David Felson, acquisition of data, analysis and interpretation of the data and revising the article. All authors gave final approval of the article to be submitted.
Competing Interest Statement: there are no conflicts of interest to disclose