Abstract
Objective
Different adipokines have been reported to play a role in the development, progression, and severity of knee osteoarthritis, but this association may be mediated by obesity. The aim of this study was to evaluate separately the associations of leptin and adiponectin with clinical severity and inflammatory markers in nonobese and obese women with knee osteoarthritis.
Design
Cross-sectional study with systematic inclusion of 115 women with symptomatic primary knee osteoarthritis. Age, physical exercise, symptoms duration, and body mass index were collected. Radiographic severity was evaluated according to Kellgren-Lawrence scale. Pain and disability were assessed by WOMAC-total, -pain, -function subscales. Two adipokines (leptin and adiponectin) and 3 inflammatory markers (TNF-α, hsCRP, and IL-6) were measured by ELISA in synovial fluid and serum.
Results
Synovial fluid adiponectin was associated with WOMAC pain, function, and total and with synovial fluid IL-6 in nonobese female knee osteoarthritis after controlling by confounders (partial correlation coefficient [PCC] = 0.395, 0.387, 0.427, and 0.649, respectively). Synovial fluid and serum leptin were significantly associated with IL-6 (PCC = 0.354) after controlling by confounders but associations with clinical severity and the rest of inflammatory markers were mitigated after control.
Conclusions
Adiponectin in synovial fluid was associated with clinical severity and local inflammatory markers in knee osteoarthritis women, while leptin relation was attenuated when controlled by confounders.
Keywords: leptin, adiponectin, knee osteoarthritis, inflammation, clinical severity
Introduction
Osteoarthritis (OA) is a chronic, degenerative joint condition, characterized by pain and disability and the knee is the most frequently involved joint. 1 The pathogenesis of OA remains unclear but age, genetics, previous trauma, and obesity are known risk factors.
Obesity is associated to low-degree systemic inflammation. Visceral tissue, and also local adipose tissue such as the infrapatellar fat pad, 2 secrete proinflammatory factors3,4 and different adipokines, with leptin and adiponectin being the most frequently studied in OA.5,6
The aim of this study was to compare the associations of synovial and serum leptin and adiponectin with clinical severity and local inflammatory factors in patients with knee OA (KOA).
Methods
Study Design and Subjects
Cross-sectional study with systematic inclusion of postmenopausal women with symptomatic primary KOA (ACR criteria) and persisting knee effusion. Symptomatic OA was defined as pain ≥4 on a 10-cm visual analogical scale. Joint effusion was confirmed by ultrasound at entry. Patients with secondary OA or with any condition that could interfere with pain perception, as well as those receiving systemic glucocorticoids therapy in the past 6 months, intra-articular glucocorticoid in the last three months or intraarticular hyaluronic acid injection in the past 6 months were excluded.
This study (ClinicalTrials.gov NCT01993342) was approved by the Local Ethical Committee at the Hospital Universitari Parc Taulí, Sabadell. All patients included were verbally informed and signed informed consent.
Assessments
Age, symptoms duration, body mass index (BMI, kg/m2), and physical exercise (never, occasional [less than 150 minutes per week], or regular) were collected. Blood analyses and joint aspiration were performed in fasting condition and at the same time of day for proper evaluation of adipokines. Synovial fluid was analyzed to ensure noninflammatory fluid (cell count <2000 cells) and absence of microcrystals. Serum and synovial samples were stored at −80°C.
Clinical severity was assessed by WOMAC index (Western Ontario and McMaster Universities Osteoarthritis Index) comprising 3 domains: pain, stiffness and physical function, and a total WOMAC score. Radiographic severity was evaluated by standing anteroposterior knee X-ray performed over the past 18 months and graded according to the Kellgren-Lawrence classification. Two rheumatologists (JC, CO) evaluated X-rays independently.
Serum and synovial levels of leptin (Biocompare, San Francisco, CA, USA). Dilution 1/100. Sensibility: <8 pg/mL, detection range: 62.5-10,000 pg/L, intra-assay: <6.3%, interassay: <7.2%) and adiponectin (eBioscience, San Diego, CA, USA). Dilution 1/1000. Sensibility: 0.01 ng/mL, detection range: 0.78-50 mg/L, coefficient intra-assay: 4.2%, coefficient interassay: 3.1%) were determined by ELISA (enzyme-linked immunosorbent assay). Synovial fluid/plasma ratios of each adipokine and leptin/adiponectin ratios in plasma and synovial fluid were performed to assess which component was associated to the outcome more consistently. As inflammatory markers tumor necrosis factor–alpha (TNF-α) and interleukin-6 (IL-6) (Luminex: HCYTOMAG-60K-03, (Merck Millipore) Dilution: direct. Sensibility: IL6: 0.9 pg/mL, TNF-α: 0.7 pg/mL, detection range: 3.2-2000 pg/L, coefficient intra-assay: IL6: 2%, TNF-α: 2.6%, coefficient interassay: IL6: 18.3%, TNF-α: 13%, and high sensitivity C-reactive protein (hsCRP) (EIA Kit, DRG Diagnostics, Marburg, Germany). Dilution 1/500. Sensibility: 0.1 mg/L. Detection range: 0.005-0.1 mg/L, coefficient intra-assay: <10%, coefficient interassay: <5%; were measured by ELISA in serum and synovial fluid.
Statistical Methods
Clinical, adipokines, and inflammatory factors data were described using nonparametric methods. Medians and interquartile ranges were used for continuous measures, while absolute and relative frequencies were applied to categorical variables.
Rounds of measurement were considered as an adjustment factor as adipokines and inflammatory factors could not be assessed at the same time in all patients due to technical reasons related to ELISA (configuration of plates used). Thus, cutoff points could not be established, but only associations.
Association analyses were carried out by separately fitting a linear model to each adipokine. For this purpose, adipokines and inflammatory factors were transformed suitably to meet the assumptions of the models. In addition, association analyses were performed after controlling by age, KOA symptom duration, BMI, physical exercise, and radiographic severity in the models. Partial correlation coefficients (PCCs) and adjusted group means derived from the models were used to measure the magnitude of the effects for continuous and categorical variables, respectively.
Five percent was set as threshold for statistical significance. All statistical analyses were conducted using R.
Results
A total of 115 patients were included ( Table 1 ). Obese patients showed shorter disease duration, did less exercise, and had worse clinical severity according to WOMAC. Serum and synovial leptin levels as well as inflammatory factors were higher in obese patients whereas serum and synovial adiponectin levels were higher in patients without obesity. No differences were observed in radiographic severity regarding obesity status.
Table 1.
Basal Characteristics, Radiographic Status, Clinical Severity Indexes, and Adipokines and Inflammatory Markers Levels in Synovial Fluid and Serum Stratified by Obesity Status. a
| Nonobese, n = 50 (43.5%) | Obese, n = 65 (56.5%) | All, n = 115 (100%) | |
|---|---|---|---|
| Age, years | 68.6 (11.6) | 68.8 (11.1) | 68.8 (11.0) |
| Evolution time, months | 55.0 (66.8) | 46.0 (75.0) | 50.0 (73.0) |
| BMI, kg/m2 | 27.1 (3.2) | 34.0 (4.5) | 30.5 (6.4) |
| Physical exercise | |||
| None | 16 (32.0%) | 37 (56.9%) | 53 (46.1%) |
| Occasional | 13 (26.0%) | 15 (23.1%) | 28 (24.3%) |
| Regular | 21 (42.0%) | 13 (20.0%) | 34 (29.6%) |
| Kellgren-Lawrence grade | |||
| 1 | 4 (8.0%) | 11 (16.9%) | 15 (13.0%) |
| 2 | 23 (46.0%) | 25 (38.5%) | 48 (41.7%) |
| 3 | 23 (46.0%) | 25 (38.5%) | 48 (41.7%) |
| 4 | 0 (0.0%) | 4 (6.2%) | 4 (3.5%) |
| WOMAC score | |||
| Pain | 10.0 (3.8) | 11.0 (3.0) | 11.0 (3.5) |
| Function | 36.0 (14.8) | 38.0 (14.0) | 37.0 (13.5) |
| Total | 48.0 (19.5) | 52.00 (21.0) | 52.0 (20.0) |
| Leptin SF, pg/mL | 33210.7 (18304.4) | 53232.3 (32845.8) | 42079.4 (29566.0) |
| Adiponectin SF, ng/mL | 2118.6 (1910.0) | 1590.1 (1003.7) | 1734.8 (1352.5) |
| TNF-α SF, pg/mL | 10.04 (8.11) | 10.16 (7.17) | 10.19 (7.97) |
| hsCRP SF, mg/mL | 0.75 (0.80) | 1.06 (0.83) | 0.91 (0.76) |
| IL-6 SF, pg/mL | 94.9 (274.1) | 122.2 (334.0) | 106.0(302.6) |
| Leptin serum, pg/mL | 28752.9 (21950.2) | 44409.2 (32622.9) | 37948.0 (28243.8) |
| Adiponectin serum, ng/mL | 18333.2 (11489.5) | 12471.6 (8681.5) | 14294.1 (9990.2) |
| TNF-α serum, pg/mL | 5.50 (1.83) | 6.00 (2.06) | 5.64 (1.95) |
| hsCRP serum, mg/mL | 4.01 (6.49) | 5.12 (5.16) | 4.92 (6.11) |
| IL-6 serum, pg/mL | 2.41 (3.29) | 2.91 (3.02) | 2.72 (3.31) |
BMI = body mass index; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; SF = synovial fluid; TNF-α = tumor necrosis factor–alpha; hsCRP = high-sensitivity C-reactive protein; IL6 = interleukin-6.
Variables are described by total frequency and percentages (categorical) or median and interquartile range (numerical). Adipokines and inflammatory factors were a priori corrected by round of measurement before summarization.
Adiponectin levels in synovial fluid showed an association with WOMAC-pain, WOMAC-function, and WOMAC-total (P < 0.05) and especially IL-6 (P < 0.001) in nonobese patients after multivariate analysis ( Table 2 ) and in patients with obesity a significant correlation with synovial fluid TNF was found (P < 0.05). However, synovial fluid leptin only showed a significant association with IL-6 (P < 0.05) after multivariate analysis. Likewise, serum leptin only showed statistically significant association with IL-6 (P < 0.01) after multivariate analysis ( Table 3 ) . No remarkable associations were found regarding serum levels of adiponectin.
Table 2.
Associations between Synovial Fluid Leptin and Adiponectin with Clinical Severity Indexes and Synovial Fluid Inflammatory Markers.
| SF Leptin | SF Adiponectin | |||||||
|---|---|---|---|---|---|---|---|---|
| Univariate | Adjusted a | Univariate | Adjusted a | |||||
| Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | |
| WOMAC pain | 0.375 [0.088, 0.605]** | 0.112 [−0.146, 0.356] | 0.258 [−0.068, 0.533] | −0.037 [−0.302, 0.233] | 0.346 [0.065, 0.576]** | −0.087 [−0.330, 0.166] | 0.395 [0.100, 0.627]** | 0.013 [−0.251, 0.275] |
| WOMAC function | 0.276 [−0.023, 0.529]* | 0.257 [0.003, 0.480]** | 0.170 [−0.158, 0.465] | 0.059 [−0.212, 0.322] | 0.350 [0.069, 0.579]** | −0.236 [−0.459, 0.015]* | 0.387 [0.090, 0.621]** | −0.114 [−0.366, 0.154] |
| WOMAC total | 0.339 [0.047, 0.578]** | 0.211 [−0.045, 0.442] | 0.244 [−0.082, 0.523] | 0.011 [−0.257, 0.278] | 0.384 [0.108, 0.604]*** | −0.210 [−0.437, 0.042] | 0.427 [0.138, 0.650]*** | −0.096 [−0.350, 0.171] |
| TNF-α SF | 0.040 [−0.288, 0.359] | 0.045 [−0.219, 0.302] | −0.044 [−0.392, 0.316] | 0.078 [−0.202, 0.346] | 0.018 [−0.299, 0.332] | 0.249 [−0.008, 0.474]* | 0.175 [−0.179, 0.489] | 0.308 [0.041, 0.533]** |
| hsCRP SF | 0.349 [0.058, 0.585]** | −0.007 [−0.261, 0.247] | 0.252 [−0.074, 0.529] | −0.058 [−0.321, 0.213] | 0.021 [−0.268, 0.306] | −0.029 [−0.276, 0.223] | 0.137 [−0.178, 0.426] | 0.020 [−0.244, 0.282] |
| IL 6 SF | 0.305 [0.005, 0.555]** | 0.002 [−0.252, 0.256] | 0.358 [0.038, 0.611]** | 0.032 [−0.238, 0.297] | 0.577 [0.344, 0.743]**** | 0.238 [−0.013, 0.460]* | 0.649 [0.423, 0.799]**** | 0.249 [−0.015, 0.481]* |
PCC = partial correlation coefficient; CI = confidence interval; SF = synovial fluid; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; TNF-α = tumor necrosis factor–alpha; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6.
The boldfaced entries are for statistically significant values.
Controlled for age, evolution time, body mass index, physical exercise, and Kellgren-Lawrence grade.
0.1 < P < 0.05, **P < 0.05, ***P < 0.01, ****P < 0.001.
Table 3.
Associations between Serum Leptin and Adiponectin with Clinical Severity Indexes and Serum Inflammatory Markers.
| Serum Leptin | Serum Adiponectin | |||||||
|---|---|---|---|---|---|---|---|---|
| Univariate | Adjusted a | Univariate | Adjusted a | |||||
| Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | |
| WOMAC pain | 0.074 [−0.217, 0.354] | 0.211 [−0.043, 0.440] | −0.056 [−0.358, 0.256] | 0.092 [−0.178, 0.348] | 0.116 [−0.177, 0.390] | −0.076 [−0.322, 0.179] | 0.184 [−0.131, 0.466] | −0.060 [−0.320, 0.208] |
| WOMAC function | 0.051 [−0.239, 0.334] | 0.335 [0.091, 0.541]*** | −0.068 [−0.368, 0.245] | 0.059 [−0.212, 0.322] | 0.004 [−0.283, 0.291] | −0.022 [−0.272, 0.231] | 0.024 [−0.286, 0.329] | 0.015 [−0.251, 0.279] |
| WOMAC total | 0.078 [−0.214, 0.357] | 0.306 [0.059, 0.518]** | −0.037 [−0.341, 0.274] | 0.129 [−0.141, 0.381] | 0.017 [−0.272, 0.303] | −0.034 [−0.283, 0.220] | 0.044 [−0.267, 0.347] | −0.009 [−0.273, 0.257] |
| TNF-α serum | 0.147 [−0.146, 0.417] | −0.099 [−0.343, 0.156] | 0.271 [−0.040, 0.534]* | −0.086 [−0.343, 0.184] | 0.120 [−0.173, 0.394] | −0.150 [−0.387, 0.106] | 0.067 [−0.246, 0.367] | −0.199 [−0.441, 0.070] |
| hsCRP serum | 0.026 [−0.264, 0.310] | 0.258 [0.006, 0.479]** | −0.178 [−0.460, 0.137] | 0.199 [−0.070, 0.441] | 0.026 [−0.263, 0.311] | 0.026 [−0.228, 0.276] | 0.104 [−0.211, 0.399] | 0.015 [−0.251, 0.280] |
| IL-6 serum | 0.059 [−0.232, 0.340] | 0.319 [0.073, 0.529]** | 0.034 [−0.276, 0.338] | 0.354 [0.098, 0.566]*** | −0.234 [−0.488, 0.057] | −0.206 [−0.435, 0.049] | −0.188 [−0.468, 0.127] | −0.252 [−0.485, 0.014]* |
WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; TNF-α = tumor necrosis factor–alpha; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6.
The boldfaced entries are for statistically significant values.
Controlled for age, evolution time, body mass index, physical exercise, and Kellgren-Lawrence scale.
0.1 < P < 0.05, **P < 0.05, ***P < 0.01, ****P < 0.001.
Synovial fluid/serum leptin ratio showed a statistically significant relation with WOMAC-pain and WOMAC-total (P < 0.05) in nonobese patients after multivariate analysis. Synovial fluid/serum adiponectin ratio showed an association with synovial fluid IL-6 in obese and nonobese patients (P < 0.001 and P < 0.01, respectively); with WOMAC-function and WOMAC-total (P < 0.05) in nonobese patients and with synovial fluid and serum TNF-α and IL-6 in obese patients (P < 0.05) ( Table 4 ) .
Table 4.
Associations between Ratio of Synovial and Serum Presence of Leptin and Adiponectin with Clinical Severity Indexes and Synovial Fluid and Serum Inflammatory Markers.
| SF/Serum Leptin | SF/Serum adiponectin | |||||||
|---|---|---|---|---|---|---|---|---|
| Univariate | Adjusted a | Univariate | Adjusted a | |||||
| Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | |
| WOMAC pain | 0.181 [−0.112, 0.445] | −0.064 [−0.308, 0.189] | 0.370 [0.057, 0.617]** | −0.121 [−0.379, 0.155] | 0.284 [0.006, 0.521]** | 0.036 [−0.212, 0.279] | 0.233 [−0.081, 0.504] | 0.088 [−0.181, 0.345] |
| WOMAC function | 0.133 [−0.160, 0.405] | −0.023 [−0.271, 0.228] | 0.304 [−0.018, 0.568]* | −0.083 [−0.346, 0.192] | 0.393 [0.129, 0.605]*** | −0.166 [−0.395, 0.084] | 0.362 [0.062, 0.603]** | −0.096 [−0.353, 0.173] |
| WOMAC total | 0.143 [−0.150, 0.413] | −0.045 [−0.292, 0.207] | 0.331 [0.012, 0.588]** | −0.115 [−0.373, 0.161] | 0.415 [0.154, 0.622]*** | −0.129 [−0.363, 0.121] | 0.384 [0.087, 0.619]** | −0.056 [−0.316, 0.213] |
| TNF-α SF | 0.063 [−0.267, 0.379] | −0.209 [−0.448, 0.057] | 0.146 [−0.220, 0.476] | −0.197 [−0.451, 0.086] | 0.165 [−0.158, 0.457] | 0.237 [−0.023, 0.467]* | 0.285 [−0.064, 0.572] | 0.275 [0.002, 0.510]** |
| hsCRP SF | 0.025 [−0.274, 0.319] | 0.065 [−0.194, 0.316] | 0.072 [−0.253, 0.383] | 0.011 [−0.260, 0.280] | 0.000 [−0.287, 0.287] | 0.086 [−0.170, 0.330] | 0.029 [−0.281, 0.334] | 0.137 [−0.133, 0.389] |
| IL-6 SF | 0.111 [−0.196, 0.398] | −0.100 [−0.348, 0.160] | 0.171 [−0.162, 0.469] | −0.106 [−0.366, 0.169] | 0.448 [0.181, 0.653]*** | 0.491 [0.273, 0.661]**** | 0.469 [0.185, 0.681]*** | 0.537 [0.317, 0.702]**** |
| TNF-α serum | −0.132 [−0.413, 0.171] | −0.066 [−0.317, 0.193] | −0.101 [−0.407, 0.226] | −0.058 [−0.323, 0.216] | 0.037 [−0.253, 0.321] | 0.243 [−0.010, 0.466]* | 0.094 [−0.220, 0.390] | 0.266 [0.001, 0.497]** |
| hsCRP serum | 0.107 [−0.196, 0.391] | 0.019 [−0.238, 0.274] | 0.202 [−0.126, 0.490] | 0.007 [−0.264, 0.277] | 0.020 [−0.269, 0.305] | −0.068 [−0.314, 0.187] | 0.064 [−0.249, 0.364] | 0.004 [−0.262, 0.269] |
| IL-6 serum | 0.054 [−0.246, 0.346] | −0.126 [−0.370, 0.134] | 0.128 [−0.199, 0.430] | −0.125 [−0.382, 0.150] | 0.036 [−0.254, 0.320] | 0.243 [−0.010, 0.466]* | 0.155 [−0.161, 0.441] | 0.308 [0.047, 0.530]** |
SF = synovial fluid; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; TNF-α = tumor necrosis factor–alpha; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6.
The boldfaced entries are for statistically significant values.
Controlled for age, evolution time, body mass index, physical exercise, and Kellgren-Lawrence scale.
0.1 < P < 0.05, **P < 0.05, ***P < 0.01, ****P < 0.001.
In the multivariate model, the synovial fluid leptin/adiponectin ratio only showed an inverse association with synovial fluid IL-6 (P < 0.01) in nonobese patients whereas serum leptin/adiponectin ratio was associated only with IL-6 in synovial fluid and serum (P < 0.01 and P < 0.001, respectively) in obese patients ( Table 5 ) .
Table 5.
Associations between Leptin and Adiponectin Ratio in Synovial Fluid and Serum with Clinical Severity Indexes and Synovial Fluid and Serum Inflammatory Markers.
| Leptin/Adiponectin SF | Leptin/Adiponectin Serum | |||||||
|---|---|---|---|---|---|---|---|---|
| Univariate | Adjusted a | Univariate | Adjusted a | |||||
| Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | Nonobese PCC [95% CI] | Obese PCC [95% CI] | |
| WOMAC pain | −0.125 [−0.398, 0.168] | 0.138 [−0.113, 0.373] | −0.284 [−0.553, 0.039]* | −0.043 [−0.307, 0.227] | 0.046 [−0.235, 0.321] | 0.198 [−0.050, 0.423] | −0.170 [−0.454, 0.145] | 0.113 [−0.157, 0.367] |
| WOMAC function | −0.167 [−0.434, 0.126] | 0.331 [0.090, 0.535]*** | −0.292 [−0.559, 0.031]* | 0.120 [−0.153, 0.376] | 0.107 [−0.177, 0.374] | 0.223 [−0.024, 0.444]* | −0.061 [−0.362, 0.251] | 0.095 [−0.175, 0.351] |
| WOMAC total | −0.173 [−0.438, 0.121] | 0.282 [0.037, 0.495]** | −0.300 [−0.566, 0.021]* | 0.071 [−0.201, 0.332] | 0.118 [−0.166, 0.384] | 0.210 [−0.038, 0.433]* | −0.056 [−0.357, 0.256] | 0.098 [−0.172, 0.354] |
| TNF-α SF | 0.047 [−0.281, 0.366] | −0.114 [−0.364, 0.151] | −0.162 [−0.489, 0.204] | −0.147 [−0.406, 0.134] | 0.117 [−0.206, 0.417] | 0.174 [−0.088, 0.414] | 0.029 [−0.318, 0.368] | 0.210 [−0.066, 0.457] |
| hsCRP SF | 0.191 [−0.113, 0.461] | 0.025 [−0.231, 0.277] | −0.006 [−0.325, 0.315] | −0.039 [−0.304, 0.231] | 0.144 [−0.149, 0.414] | 0.058 [−0.196, 0.306] | −0.011 [−0.317, 0.298] | 0.076 [−0.193, 0.335] |
| IL-6 SF | −0.373 [−0.606, −0.082]** | −0.166 [−0.403, 0.092] | −0.475 [−0.692, −0.178]*** | −0.186 [−0.432, 0.086] | −0.030 [−0.318, 0.262] | 0.266 [0.015, 0.485]** | −0.057 [−0.362, 0.259] | 0.362 [0.106, 0.572]*** |
| TNF-α serum | −0.092 [−0.379, 0.210] | −0.179 [−0.416, 0.081] | −0.030 [−0.346, 0.293] | −0.174 [−0.424, 0.102] | 0.009 [−0.279, 0.295] | 0.036 [−0.218, 0.285] | 0.125 [−0.189, 0.417] | 0.099 [−0.171, 0.355] |
| hsCRP serum | 0.054 [−0.247, 0.345] | 0.215 [−0.044, 0.446] | −0.194 [−0.484, 0.134] | 0.125 [−0.150, 0.382] | −0.002 [−0.289, 0.285] | 0.171 [−0.084, 0.406] | −0.190 [−0.470, 0.125] | 0.128 [−0.142, 0.380] |
| IL-6 serum | 0.225 [−0.077, 0.489] | 0.114 [−0.146, 0.360] | 0.091 [−0.235, 0.399] | 0.117 [−0.159, 0.375] | 0.199 [−0.094, 0.459] | 0.386 [0.149, 0.581]*** | 0.159 [−0.156, 0.445] | 0.450 [0.210, 0.639]**** |
SF = synovial fluid; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; TNF-α = tumor necrosis factor–alpha; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6.
The boldfaced entries are for statistically significant values.
Controlled for age, evolution time, body mass index, physical exercise, and Kellgren-Lawrence scale.
0.1 < P < 0.05, **P < 0.05, ***P < 0.01, ****P < 0.001.
Discussion
In this study, synovial fluid adiponectin was related to clinical severity and synovial fluid IL-6 in non-obese patients, while in obese patients it was only associated to synovial fluid TNF-α. The ratio synovial fluid/serum adiponectin was in line with these results, pointing to the importance of the synovial fluid component as local inflammatory marker. Although this effect appeared more clearly in nonobese patients, the ratio suggested a mild effect in obese patients. The serum leptin/adiponectin ratio was also in accordance with the local inflammatory effect of adiponectin as evidenced by the association with synovial fluid IL-6 in nonobese patients. Therefore, these results support a local proinflammatory effect of synovial adiponectin and an association with clinical severity in nonobese women with KOA.
A prior study in this cohort of patients analyzing adipokines in synovial fluid controlled by inflammatory markers and anthropometric measurements revealed an association between adiponectin and pain. 7 Previous studies have linked clinical severity with adiponectin in plasma and synovial fluid and with the adiponectin/leptin ratio,8-10 although these studies had differences regarding presence of joint effusion, gender homogeneity, and radiographic stage compared with our sample. However, one of these studies pointed to a relation of adiponectin with synovial IL-6 in agreement with our data. 9 The association between adiponectin and IL-6 in synovial fluid has been previously reported 11 but was not evaluated separately according to obesity status.12,13 In our study, all patients were women and this fact could increase the capacity to find associations as gender-specific correlations between adiponectin and pain have been communicated. 14
Synovial fluid and serum leptin associations with clinical severity and inflammatory markers in obese patients were attenuated after multivariate analysis, except for serum leptin and serum IL-6, which could represent the systemic inflammatory effect of this adipokine. In nonobese patients, the only association that remained after adjustments was between synovial fluid leptin and synovial fluid IL-6, pointing to an inflammatory local effect. In accordance with our results, a previous study in patients with KOA 15 linked adiposity and serum leptin to clinical status in obese patients, although no adjustment by confounders were performed in that previous work.
The synovial/serum leptin ratio was associated with clinical severity in nonobese patients, in accordance with the association observed with synovial fluid leptin. The leptin/adiponectin ratio in synovial fluid revealed a greater effect of synovial adiponectin as a local inflammatory marker.
The main limitation of our study arises from its cross-sectional design. Accordingly, conclusions can only be drawn in terms of associations. Selection bias toward more severe disease could exist as all patients were referred to our Rheumatology Unit for specialist care. Additionally, subjects in the nonobese group were predominantly overweight and therefore our results cannot be applied to normal-weight patients.
A strength of this work is the homogeneity of the study sample as all patients were women with significant symptomatic KOA with joint effusion and had predominantly low-to-moderate radiographic stage. This homogeneity is relevant as results can be applied to a more definite phenotype and might increase the statistical power to detect associations of a moderate magnitude. To our knowledge this is the first time synovial fluid and serum samples have been evaluated in obese and nonobese separately.
Replication is warranted, especially in other groups of patients, such as men or in low- or normal-weight patients.
In conclusion, synovial fluid adiponectin in women with KOA was associated with clinical severity and synovial fluid IL-6. Serum leptin and adiponectin and synovial leptin showed no significant association with clinical severity and local inflammation in this population.
Footnotes
Author Contributions: CO and JC: participated in the conception and design of the study, acquisition of data and analysis and interpretation of the results; drafting the article and revising critically for important intellectual content and final approval of the version to be submitted. AB: participated in the design of the study, in the analysis and interpretation of the results, revising the article critically for important intellectual content and final approval of the version to be submitted. NA: participated in the conception of the study, analysis of synovial fluid, revising the article critically for important intellectual content and final approval of the version to be submitted. MG and CG: participated in acquisition data, interpretation of the results, revising the article critically for important intellectual content and final approval of the version to be submitted. AC: participated in the conception of the study, acquisition data, revising the article critically for important intellectual content and final approval of the version to be submitted. MA and MLL: participated in acquisition data, results analysis and interpretation and final approval of the version to be submitted. JG: participated in the conception and design of the study, interpretation of the results, revising the article critically for important intellectual content and final approval of the version to be submitted.
Acknowledgment and Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the grant for investigational osteoarthritis studies from “Societat Catalana de Reumatologia” years 2015 and 2016 and the grant for investigational project without public grant of “Sociedad Española de Reumatología” 2017.
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical Approval: This study was approved by the Local Ethical Committee at the Hospital Universitari Parc Taulí, Sabadell.
Informed Consent: All patients included were verbally informed and signed informed consent.
Trial Registration: ClinicalTrials.gov identifier: NCT01993342.
ORCID iD: Joan Calvet 
https://orcid.org/0000-0002-0888-5152
References
- 1. Cross M, Smith E, Hoy D, Nolte S, Ackerman I, Fransen M, et al. The global burden of hip and knee osteoarthritis: estimates from the global burden of disease 2010 study. Ann Rheum Dis. 2014;73(7):1323-30. [DOI] [PubMed] [Google Scholar]
- 2. Conde J, Scotece M, López V, Abella V, Hermida M, Pino J, et al. Differential expression of adipokines in infrapatellar fat pad (IPFP) and synovium of osteoarthritis patients and healthy individuals. Ann Rheum Dis. 2014;73(3):631-3. [DOI] [PubMed] [Google Scholar]
- 3. Wang T, He C. Pro-inflammatory cytokines: the link between obesity and osteoarthritis. Cytokine Growth Factor Rev. 2018;44:38-50. [DOI] [PubMed] [Google Scholar]
- 4. Stannus OP, Jones G, Blizzard L, Cicuttini FM, Ding C. Associations between serum levels of inflammatory markers and change in knee pain over 5 years in older adults: a prospective cohort study. Ann Rheum Dis. 2013;72(4):535-40. [DOI] [PubMed] [Google Scholar]
- 5. Tu C, He J, Wu B, Wang W, Li Z. An extensive review regarding the adipokines in the pathogenesis and progression of osteoarthritis. Cytokine. 2019;113:1-12. [DOI] [PubMed] [Google Scholar]
- 6. de Boer TN, van Spil WE, Huisman AM, Polak AA, Bijlsma JW, Lafeber FP, et al. Serum adipokines in osteoarthritis; comparison with controls and relationship with local parameters of synovial inflammation and cartilage damage. Osteoarthritis Cartilage. 2012;20(8):846-53. [DOI] [PubMed] [Google Scholar]
- 7. Calvet J, Orellana C, Giménez NA, Berenguer-Llergo A, Caixàs A, García-Manrique M, et al. Differential involvement of synovial adipokines in pain and physical function in female patients with knee osteoarthritis. A cross-sectional study. Osteoarthritis Cartilage. 2018;26(2):276-84. [DOI] [PubMed] [Google Scholar]
- 8. Cuzdan Coskun N, Ay S, Evcik FD, Oztuna D. Adiponectin: is it a biomarker for assessing the disease severity in knee osteoarthritis patients? Int J Rheum Dis. 2017;20(12):1942-9. [DOI] [PubMed] [Google Scholar]
- 9. Bas S, Finckh A, Puskas GJ, Suva D, Hoffmeyer P, Gabay C, et al. Adipokines correlate with pain in lower limb osteoarthritis: different associations in hip and knee. Int Orthop. 2014;38(12):2577-83. [DOI] [PubMed] [Google Scholar]
- 10. Staikos C, Ververidis A, Drosos G, Manolopoulos VG, Verettas DA, Tavridou A. The association of adipokine levels in plasma and synovial fluid with the severity of knee osteoarthritis. Rheumatology (Oxford). 2013;52(6):1077-83. [DOI] [PubMed] [Google Scholar]
- 11. Gross JB, Guillaume C, Gégout-Pottie P, Mainard D, Presle N. Synovial fluid levels of adipokines in osteoarthritis: association with local factors of inflammation and cartilage maintenance. Biomed Mater Eng. 2014;24(1 Suppl):17-25. [DOI] [PubMed] [Google Scholar]
- 12. Siqueira MB, Frangiamore S, Klika AK, Gajewski N, Barsoum WK, Higuera CA. Comparison of synovial fluid cytokine levels between traumatic knee injury and end-stage osteoarthritis. J Knee Surg. 2017;30(2):128-133. [DOI] [PubMed] [Google Scholar]
- 13. Orita S, Koshi T, Mitsuka T, Miyagi M, Inoue G, Arai G, et al. Associations between proinflammatory cytokines in the synovial fluid and radiographic grading and pain-related scores in 47 consecutive patients with osteoarthritis of the knee. BMC Musculoskelet Disord. 2011;12:144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Perruccio AV, Mahomed NN, Chandran V, Gandhi R. Plasma adipokine levels and their association with overall burden of painful joints among individuals with hip and knee osteoarthritis. J Rheumatol. 2014;41(2):334-7. [DOI] [PubMed] [Google Scholar]
- 15. Richter M, Trzeciak T, Rybka JD, Suchorska W, Augustyniak E, Lach M, et al. Correlations between serum adipocytokine concentrations, disease stage, radiological status and total body fat content in the patients with primary knee osteoarthritis. Int Orthop. 2017;41(5):983-9. [DOI] [PubMed] [Google Scholar]