Clin Rheumatol
DOI 10.1007/s10067-014-2708-x
ORIGINAL ARTICLE
Relationships of common polymorphisms in IL-6, IL-1A,
and IL-1B genes with susceptibility to osteoarthritis:
a meta-analysis
Hao Cai & Huan-Jian Sun & You-Hua Wang & Zhe Zhang
Received: 22 April 2014 / Revised: 9 May 2014 / Accepted: 30 May 2014
# Clinical Rheumatology 2014
Abstract Observational and experimental studies have ar-
rived at inconsistent conclusions about whether common
polymorphisms in IL-6, IL-1A, and IL-1B genes are associated
with an increased risk of osteoarthritis (OA). Therefore, we
undertook a comprehensive meta-analysis to more systemat-
ically summarize the relationships of IL-6, IL-1A, and IL-1B
genetic polymorphisms with susceptibility to OA. We
screened the PubMed, Embase, Web of Science, Cochrane
Library, CISCOM, CINAHL, Google Scholar, China
BioMedicine (CBM), and China National Knowledge
Infrastructure (CNKI) databases up to 31 March 2014. We
used STATA software to analyze statistical data. Odds ratios
(ORs) and their corresponding 95 % confidence intervals
(95 % CIs) were calculated. Seventeen independent case–
control studies were included in this meta-analysis with a total
number of 7,491 subjects, comprised of 3,293 OA patients
and 4,729 healthy controls. Our results indicate that IL-6, IL-
1A, and IL-1B genetic polymorphisms are statistically corre-
lated with an increased risk of OA under the allele and
dominant models. According to a subgroup analysis based
on disease, a higher frequency of IL-6 genetic polymorphisms
was observed among knee OA and hand OA patients, but not
among hip OA and DIP OA patients. A higher frequency of
IL-1A genetic polymorphisms were found among hip OA
patients, hand OA, hip OA and DIP OA patients.
Furthermore, we observed a higher IL-1B polymorphism fre-
quency among knee OA and hip OA patients, but not among
hand OA patients. Our findings provide evidence that IL-6,
H. Cai (*) : H.<J. Sun : Z. Zhang
Department of Orthopedics, Third People’s Hospital of Yancheng,
Xindu Road No.606, Yancheng 224000, People’s Republic of China
e-mail: caihao422@126.com
Y.<H. Wang
Department of Orthopedics, The Affiliated Hospital of Nantong
University, Nantong 226000, People’s Republic of China
IL-1A, and IL-1B genetic polymorphisms may be correlated
with susceptibility to OA.
Keywords IL-6 . IL-1A . IL-1B . Osteoarthritis .
Polymorphism
Introduction
Osteoarthritis (OA), the most common form of arthritis, is a
chronic and progressive disease of the joints that is a substan-
tial medical, social, and economic burden [1]. Generally
speaking, the clinical symptoms of OA may be the degener-
ation of articular cartilage and formation of abnormal bone
[2]. As the greatest cause of disability, OA has a higher
incidence and prevalence in middle- and older-aged individ-
uals [3]. The World Health Organization Scientific Group on
Rheumatic Diseases estimates that among those people who
are 60 or older, the incidence rate of OA is up to 1 in 10 [4]. In
addition, different large joints represent differences in risks,
and women have a higher risk of developing knee OA and hip
OA than men [5]. This complex multifactorial disease is
widely accepted to be induced by an interaction of environ-
mental and genetic factors [6, 7]. Multiple risk factors includ-
ing age, sex, obesity, ethnicity, and occupation, alongside
genetic components may be associated with susceptibility to
OA [6]. Despite these well-recognized environmental factors,
genetic variations have also been found to be strong determi-
nants, and several genes have been extensively documented
by accumulating studies [8].
Interleukins (ILs) are a group of secreted proteins and
signaling molecules that are capable of promoting cell growth,
differentiation, and functional activation [9]. Generally, the
majority of ILs are synthesized by helper CD4 T lymphocytes
as well as through monocytes, macrophages, and endothelial
cells [10]. To be specific, macrophages, which play crucial

roles in both immunity and inflammation, are responsible for
the production of IL-1 and IL-6 [9].
IL-1 was first identified as a protein that could induce fever
and was consequently called human leukocytic pyrogen. IL-1
is now recognized as consisting of two major proteins, IL-1A
and IL-1B [11]. IL-1A is considered to possess metabolic,
physiological, and hematopoietic activities and to play an
important role in the regulation of immune responses [10].
IL-1B is also a crucial mediator of the inflammatory response
and participates in a variety of cellular activities, including cell
proliferation, differentiation, and apoptosis [11]. In recent
years, IL-1A and IL-1B have been suggested to be involved
in the development of OA [5, 12]. There is evidence illustrating
that inflammatory components may be implicated in OA de-
velopment, and IL-1A and IL-1B have also been found in the
synovial fluid and cartilage of OA patients [13]. Since IL-1A
and IL-1B may facilitate cartilage degradation and contribute to
producing other substances, which also serve to accelerate
cartilage breakdown, IL-1A and IL-1B loci have been regarded
as biologically possible candidate genes for OA [14]. The
human IL-1A gene is located on the long arm of chromosome
2 at band q13, in the same chromosomal region as the IL-1B
gene [15]. Genetic polymorphisms in the IL-1A and IL-1B
genes may have impacts on the expression of IL-1A and IL-
1B proteins, which may give rise to the stimulation of cartilage
breakdown, thereby conducing to the pathogenesis of OA [16].
On the other hand, IL-6, which is supposed to be one of the
main factors in joint destruction, is a pleiotropic pro-
inflammatory cytokine that is obviously elevated during tissue
inflammation [17]. Large quantities of evidence have shown
that IL-6 levels in the serum and synovial fluid is up-regulated
in OA patients [18]. Recently, a large number of epidemio-
logical studies have demonstrated that IL-6 genetic polymor-
phisms may be significantly correlated with the pathogenesis
of OA [19, 20]. The human IL-6 gene is located on chromo-
some 7p21–24 and contains five exons and four introns, with
an upstream promoter comprised of 303 bp [21]. Genetic
variants in the promoter region of the IL-6 gene may cause
alterations in transcription that subsequently affect circulating
levels of human recombinant IL-6, which is suspected to be
able to reinforce human recombinant IL-1β-induced proteo-
glycan degradation and suppress chondrocyte proliferation,
and thus promote the progress of OA [8].
Several studies have supported the notion that polymorphic
variants in the IL-1A, IL-1B, and IL-6 genes might be closely
related with OA pathogenesis; these polymorphisms include
rs1800587 C > T in IL-1A, rs16944 C > T and rs1143634 C >
T in IL-1B, and rs1800795 G > C in IL-6 [5, 6, 12, 22].
Nevertheless, inconsistent results have also been reported by
other studies [19, 22]. Therefore, we performed the current
meta-analysis to systematically evaluate the correlation of
genetic polymorphisms in the IL-1A, IL-1B, and IL-6 genes
with the pathogenesis of OA.
Clin Rheumatol
Materials and methods
Search strategy
We conducted a comprehensive search for relevant studies pub-
lished before 31 March 2014 on the PubMed, Embase, Web of
Science, Cochrane Library, CISCOM, CINAHL, Google
Scholar, China BioMedicine (CBM), and China National
Knowledge Infrastructure (CNKI) databases. We used the fol-
lowing series of keywords and MeSH terms: (“Osteoarthritis,
Spine” or “Osteoarthritis, Knee” or “Osteoarthritis, Hip” or
“Osteoarthritis” or “knee osteoarthritis” or “spine osteoarthritis”
or “hip osteoarthritis” or “spinal osteoarthritis” or “lumbar oste-
oarthritis” or “coxarthrosis”), (“Interleukins” or “IL” or
“Interleukins”), and (“Polymorphism, Genetic” or “polymor-
phism” or “polymorphisms” or “variants” or “SNP” or “muta-
tion” or “genetic variants”). No language restrictions were used.
A manual search of reference lists was also performed to identify
other potential studies.
Selection criteria
The studies that met the following criteria were enrolled in the
current meta-analysis: (1) nested case–control studies or case–
control studies concentrating on the polymorphisms of inter-
leukins in OA; (2) studies including a minimum number of 30
cases; (3) patients who were diagnosed with distal interpha-
langeal (DIP) OA, hand OA, and hip and knee OA were
assessed clinically and radiographically based on physical
examination followed by roentgenogram; patients who were
diagnosed with any other metabolic, endocrinological or
bone-related disorders, or had underwent hormone replace-
ment therapy were excluded; (4) sufficient data was published
on genotype and mutation frequencies. Studies were excluded
if they did not meet all of these inclusion criteria. If more than
one study was published using the same case series, either the
study with the largest sample size or the one most recently
published was enrolled. Any disagreements were resolved by
discussions and subsequent consensus.
Data extraction
Data was extracted from eligible studies by two authors inde-
pendently using a standardized form. The following informa-
tion was collected: surname of first author, year of publication,
source of publication, country of origin, ethnicity, language of
publication, study type, total number of subjects or samples,
source of subjects or samples, pathological subtype, types of
DNA samples, detection method of polymorphism, and geno-
type and mutation frequencies. Discrepancies in data collected
from the eligible studies were resolved through discussion and
careful reexamination of the full text by the authors to avoid
conflicting evaluations.
Clin Rheumatol

Fig. 1 Flow chart shows study Articles identified through Additional articles identified

Identification
selection procedure. Seventeen electronic database searching through a manual search
case–control studies were (N = 94) (N = 1)
included in this meta-analysis

Articles reviewed for duplicates

(N = 95)

Screening
Articles after duplicates removed
(N = 93)
Studies were excluded, due to:
(N = 6) Letters, reviews, meta-analysis
(N = 11) Not human studies
(N = 15) Not related to research topics

Full-text articles assessed

for eligibility
(N = 61)
Eligibility

Studies were excluded, due to:

(N = 8) Not case-control study
(N = 13) Not relevant to IL-6 and IL-1A
IL-1B gene
(N = 21) Not relevant to osteoarthritis
Studies included in
qualitative synthesis
(N = 19)
Included

Studies included in quantitative

synthesis (meta-analysis)
(N = 17)

Quality assessment studies using Cochran’s Q-statistic, which is considered sig-

Two authors independently assessed the quality of included
studies according the Strengthening the Reporting of Genetic
Association Studies (STREGA) quality score systems and the
Newcastle-Ottawa Scale (NOS) [23, 24]. The STREGA scale
consisting of 22 assessment items correlated to quality ap-
praisal, with scores ranging from 0 to 22. The included studies
were classified into three levels based on their scores: low
quality (0–12), moderate quality (13–17), and high quality
(18–22). The NOS criteria use a “star” rating system to judge
methodological quality, which is based on three aspects of a
study as follows: selection, comparability, and exposure.
Scores range from 0 (worst) to 9 stars (best), with a score
higher than 7 indicating that a study’s methodological quality
was generally good. Any disagreements over STREGA and
nificant at P<0.05 [25]. The I2 test was also used to quantify
heterogeneity (ranges from 0 to 100 %) [26]. The random-
effects model (DerSimonian-Laird method) was used wherever
significant heterogeneity was found, indicated by a P<0.05 or
I2 >50 %. Wherever no statistical heterogeneity was found, we
used a fixed effects model (Mantel-Haenszel method). In order
to explore potential sources of heterogeneity, subgroup analy-
ses were performed based on disease and genotyping method.
To evaluate the influence of single studies on the overall
estimate, we conducted a sensitivity analysis by omitting each
Pubmed database
20
All database
15
Number of articles

NOS score evaluations for the included studies were resolved

10
through a comprehensive reassessment by the authors.

5
Statistical analysis

Odds ratios (ORs) and their 95 % confidence intervals (95 % 0

2001~2002 2003~2004 2005~2006 2007~2008 2009~2010 2011~2012 2013~2014

CIs) were calculated using either a fixed or random-effects Publication year

model. The significance of pooled estimates was made using Fig. 2 The distribution of the number of topic-related literature in the
the Z test. We estimated the degree of heterogeneity among electronic database over the last decade
Table 1 Characteristics of included studies in this meta-analysis

First author Year Ethnicity Disease Number Gender (M/F) Age (years) Genotyping method Gene NOS score STREGA score

Case Control Case Control Case Control

Shen CK [31] 2013 Asians Knee OA 120 130 32/88 39/91 65.8±6.8 64.1±7.1 PCR-RFLP IL-1β 8 21
Wu X [33] 2013 Caucasians Knee OA 88 66 29/59 25/41 64.6±10.6 64.5±10.6 SBE IL-1α 7 20
IL-1β
Kaarvatn MH [5] 2013 Caucasians Knee OA 240 531 66/174 398/133 69.7±7.2 42.3±11.9 TaqMan Assay IL-1α 8 21
Hip OA 260 531 85/175 398/133 67.8±9.6 42.3±11.9 TaqMan Assay IL-1α 8 21
Chen Y [29] 2012 Asians Knee OA 82 102 35/47 42/60 61 59 PCR-RFLP IL-1β 7 20
Jotanovic Z [6] 2012 Caucasians Knee OA 238 458 65/173 – 69.7±7.2 – TaqMan Assay IL-1β 8 21
Solovieva S [32] 2009 Caucasians DIP OA 335 207 – – 55.7±4.8 52.8±5.2 PCR-RFLP IL-1α/IL-1β 8 21
Ni H [1] 2009 Asians Knee OA 453 487 130/323 173/314 59.0±11.7 56.9±11.6 PCR-RFLP IL-1β 8 21
Sezgin M [16] 2007 Caucasians Knee OA 107 67 26/81 21/47 61.3±8.9 51.5±8.9 PCR-RFLP IL-1α
Moxley G [12] 2007 Caucasians Hand OA 64 48 – – – – TaqMan Assay IL-1α/IL-1β 5 18
Chapman K [28] 2006 Caucasians Hip OA 370 544 142/228 339/205 – – PCR-RFLP IL-1β 8 21
Stern AG [14] 2003 Caucasians Hand OA 68 51 14/54 45/6 – – TaqMan Assay IL-1α/IL-1β 6 19
Loughlin J [30] 2002 Caucasians Hip OA/Knee OA 557 557 215/342 342/215 – – PCR-RFLP IL-1α/IL-1β 8 21
Kolundzic R [19] 2011 Caucasians DDH-OA 28 20 5/23 9/11 44 (34~57) 21 (20~35) PCR-RFLP IL-6
Honsawek S [8] 2011 Asians KOA 115 100 18/97 15/85 65.5±1.1 63.8±2.1 PCR-RFLP IL-6 8 20
Kamarainen OP [22] 2008 Caucasians DIPOA 48 487 – – 57.7±4.2 53.6±5.2 TaqMan Assay IL-6 5 18
Pola E [20] 2005 Caucasians HOA 75 107 30/45 55/52 70.4±7.9 71.7±10.8 PCR-RFLP IL-6 6 19
Moos V [34] 2000 Caucasians Knee OA 45 236 – – – – PCR-RFLP IL-1β 5 18

M male, F female, NOS the Newcastle-Ottawa Scale, STREGA strengthening the reporting of genetic association studies
Clin Rheumatol
Clin Rheumatol

Table 2 Meta-analysis of the relationships of IL-6, IL-1A and IL-1B with osteoarthritis susceptibility

M allele vs. W WM + MM vs. WW MM vs. WW + WM MM vs. WW MM vs. WM

(Allele model) (Dominant model) (Recessive model) (Homozygous model) (Heterozygous model)

OR 95 % CI P OR 95 % CI P OR 95 % CI P OR 95 % CI P OR 95 % CI P

IL-6 1.59 1.21–2.10 0.001 1.91 1.36–2.68 <0.001 1.44 0.78–2.65 0.244 1.88 0.97–3.64 0.060 1.15 0.60–2.91 0.677
Disease
Hip OA 1.23 0.41–3.72 0.711 1.00 0.27–3.76 1.000 2.28 0.22–23.68 0.490 2.14 0.20–22.65 0.526 3.00 0.21–42.62 0.417
Knee OA 2.91 1.47–5.76 0.002 3.34 1.63–6.87 0.001 – – – – – – – – –
DIP OA 1.32 0.89–1.95 0.163 1.67 0.96–2.89 0.069 1.00 0.41–2.46 0.992 1.44 0.52–3.95 0.481 0.81 0.32–2.04 0.655
Hand OA 1.63 1.04–2.55 0.033 1.73 0.96–3.13 0.068 1.92 0.79–4.71 0.152 2.33 0.91–5.95 0.078 1.48 0.56–3.89 0.427
Genotype method
PCR-RFLP 1.87 1.22–2.87 0.004 2.03 1.13–3.65 0.018 1.97 0.85–4.54 0.113 2.30 0.96–5.50 0.061 1.61 0.65–3.99 0.306
Non-PCR- 1.32 0.89–1.95 0.163 1.67 0.96–2.89 0.069 1.00 0.41–2.46 0.992 1.44 0.52–3.95 0.481 0.81 0.32–2.04 0.655
RFLP
IL-1A 1.12 1.01–1.25 0.032 1.19 1.03–1.39 0.021 0.25 0.21–0.30 <0.001 0.27 0.22–0.33 <0.001 0.05 0.04–0.07 <0.001
Disease
Hip OA 1.31 1.11–1.56 0.002 1.48 1.18–1.86 0.001 0.29 0.21–0.40 <0.001 0.34 0.24–0.48 <0.001 0.06 0.04–0.09 <0.001
Knee OA 1.02 0.88–1.19 0.812 1.06 0.87–1.29 0.551 0.22 0.16–0.29 <0.001 0.23 0.17–0.31 <0.001 0.05 0.03–0.07 <0.001
DIP OA 1.04 0.80–1.34 0.789 0.98 0.69–1.39 0.906 0.27 0.18–0.41 <0.001 0.26 0.17–0.41 <0.001 0.06 0.03–0.11 <0.001
Hand OA 1.05 0.59–1.89 0.865 1.15 0.55–2.38 0.711 0.11 0.03–0.39 0.001 0.12 0.03–0.45 0.002 0.01 0.00–0.08 <0.001
Genotype method
PCR-RFLP 1.11 0.95–1.30 0.181 1.24 1.01–1.52 0.041 0.23 0.17–0.32 <0.001 0.26 0.18–0.38 <0.001 0.05 0.03–0.08 <0.001
Non-PCR- 1.16 0.96–1.40 0.124 1.20 0.92–1.56 0.176 0.26 0.20–0.33 <0.001 0.27 0.21–0.34 <0.001 0.050. 0.04–0.07 <0.001
RFLP
IL-1B 1.27 1.20–1.35 <0.001 1.32 1.21–1.43 <0.001 1.45 1.29–1.64 <0.001 1.63 1.44–1.84 <0.001 1.31 1.14–1.50 <0.001
Disease
Knee OA 1.24 1.14–1.34 <0.001 1.25 1.12–1.39 <0.001 1.47 1.24–1.73 <0.001 1.63 1.38–1.93 <0.001 1.36 1.15–1.62 <0.001
Hip OA 1.36 1.15–1.61 <0.001 1.40 1.13–1.72 0.002 1.71 1.38–2.13 <0.001 1.91 1.52–2.39 <0.001 1.49 1.19–1.87 0.001
DIP OA 1.28 1.03–1.59 0.026 1.48 0.98–2.23 0.062 1.02 0.67–1.54 0.940 1.24 0.83–1.84 0.295 0.81 0.40–1.64 0.553
Hand OA 1.18 0.96–1.45 0.106 1.23 0.94–1.60 0.136 1.19 0.79–1.79 0.398 1.26 0.82–1.94 0.289 1.08 0.68–1.72 0.755
Genotype method
PCR-RFLP 1.30 1.20–1.41 <0.001 1.36 1.22–1.52 <0.001 1.49 1.26–1.76 <0.001 1.68 1.44–1.96 <0.001 1.33 1.10–1.60 0.003
Non-PCR- 1.20 1.08–1.34 0.001 1.21 1.05–1.38 0.008 1.41 1.10–1.80 0.006 1.51 1.17–1.96 0.002 1.28 0.99–1.67 0.061
RFLP

W wild allele, M mutant allele, WW wild homozygote, WM heterozygote, MM mutant homozygote, OR odds ratio, 95 % CI 95 % confidence interval

study in turn. To investigate whether publication bias might
influence the validity of the estimates, funnel plots were con-
structed. The symmetry of the funnel plot was further evaluat-
ed by Egger’s linear regression test [27]. All tests were two-
sided, and a P value of <0.05 was considered statistically
significant. All analyses were calculated using the STATA
software, version 12.0 (Stata Corp, College Station, TX, USA).
Results
Baseline characteristics of included studies
Initially, we identified a total of 95 articles associated with the
searched keywords, 34 of which were excluded after
reviewing their titles and key words, and 44 of which were
then excluded after the review of the abstract and full text. At
last, 17 studies met our inclusion criteria for this meta-analysis
[1, 5, 6, 8, 12, 14, 16, 19, 20, 22, 28–34]. The flow chart of the
study selection process is shown in Fig. 1. Publication years
ranged from 2000 to 2013. Figure 2 shows the distribution of
the number of topic-related literatures in the electronic data-
base over the last decade. A total number of 7,491 subjects
were involved in this meta-analysis, including 3,293 OA cases
and 4,729 healthy controls. Peripheral blood samples were
used for the determination of interleukin polymorphisms.
Genotype methods include polymerase chain reaction-
restriction fragment length polymorphism (PCR-RFLP),
single-base extension (SBE), and TaqMan assay. Hardy-
Weinberg Equilibrium (HWE) tests were performed in all
included studies. None of the studies deviated from the
 Clin Rheumatol

IL-6 IL-6
Fig. 3 Forest plots for the Included study
(M allele versus W allele)
OR (95%CI) Weight% Included study
(WM+MM versus WW)
OR (95%CI) Weight%
relationships of IL-6, IL-1A, and Kolundzic R (2011) 1.23 (0.41, 3.72) 6.04 Kolundzic R (2011) 1.00 (0.27, 3.76) 6.58

IL-1B genetic polymorphisms Honsawek S (2011) 2.91 (1.47, 5.76) 15.36 Honsawek S (2011) 3.34 (1.63, 6.87) 22.26

with susceptibility to Kamarainen OP−a (2008) 1.28 (0.84, 1.96) 37.11 Kamarainen OP−a (2008) 1.71 (0.88, 3.31) 26.33

osteoarthritis under the allele and Kamarainen OP−b (2008) 1.55 (0.59, 4.05) 7.90 Kamarainen OP−b (2008) 1.58 (0.59,
) 4.25) 11.78

dominant models Pola E (2005) 1.63 (1.04, 2.55) 33.59 Pola E (2005) 1.73 (0.96, 3.13) 33.05

Heterogeneity test (I2 = 5.3%, P = 0.377) 1.59 (1.21, 2.10) 100.00 Heterogeneity test (I2 = 0.00%, P = 0.465) 1.91 (1.36, 2.68) 100.00
Z test (Z = 3.33, P = 0.001) Z test (Z = 3.72, P < 0.001)

0.174 1 5.76 0.146 1 6.87

IL-1A IL-1A
(M allele versus W allele) (WM+MM versus WW)
Included study OR (95%CI) Weight% Included study OR (95%CI) Weight%

Kaarvatn MH−a (2013) 1.22 (0.97, 1.54) 19.16 Kaarvatn MH−a (2013) 1.40 (1.03, 1.91) 18.58

Kaarvatn MH−b (2013) 1.02 (0.81, 1.29) 19.61 Kaarvatn MH−b (2013) 1.12 (0.83, 1.51) 19.81

Solovieva S (2009) 1.04 (0.80, 1.34) 15.59 Solovieva S (2009) 0.98 (0.69, 1.39) 15.33

Sezgin M (2007) 1.60 (0.95, 2.70) 4.02 Sezgin M (2007) 1.91 (1.02, 3.60) 5.31

Stern AG (2003) 1.05 (0.59, 1.89) 3.21 Stern AG (2003) 1.15 (0.55, 2.38) 4.06

Loughlin J−a (2002) 1.01 (0.83, 1.24) 24.99 Loughlin J−a (2002) 1.02 (0.78, 1.32) 24.06

Loughlin J−b (2002) 1.38 (1.04, 1.82) 13.41 Loughlin J−b (2002) 1.48 (1.00, 2.18) 12.86

Heterogeneity test (I2 = 5.2%, P = 0.388) 1.12 (1.01, 1.25) 100.00 Heterogeneity test (I2 = 16.1%, P = 0.307) 1.19 (1.03, 1.39) 100.00
Z test (Z = 2.15, P = 0.032) Z test (Z = 2.30, P = 0.021)

0.37 1 2.7 0.278 1 3.6

IL-1B
(M allele versus W allele)
Included study OR (95%CI) Weight%
Shen CK−a (2013) 1.05 (0.74, 1.50) 2.53
Shen CK−b (2013) 1.11 (0.78, 1.58) 2.54
Wu X−a (2013) 1.48 (0.80, 2.72) 0.92
Wu X−b (2013) 1.14 (0.71, 1.80) 1.55
Wu X−c (2013) 1.06 (0.65, 1.73) 1.42
Kaarvatn MH−a (2013) 1.07 (0.83, 1.38) 4.34
Kaarvatn MH−b (2013) 1.09 (0.85, 1.39) 4.54
Chen Y (2012) 1.22 (0.72, 2.06 1.22
Jotanovic Z (2012) 1.50 (1.18, 1.90) 4.87
Solovieva S−a (2009) 1.09 (0.82, 1.45) 3.60
Solovieva S−b (2009) 1.61 (1.19, 2.18) 3.26
Solovieva S−c (2009) 1.22 (0.95, 1.58) 4.30
Ni H−a (2009) 1.19 (0.99, 1.43) 7.02
Ni H−b (2009) 1.29 (1.03, 1.60) 5.39
Ni H−c (2009) 1.36 (0.99, 1.88) 2.94
Sezgin M (2007) 1.04 (0.61, 1.76 1.21
Moxley G−a (2007) 1.10 (0.56, 2.14) 0.78
Moxley G−b (2007) 1.35 (0.78, 2.34) 1.13
Moxley G−c (2007) 1.07 (0.61, 1.90) 1.06
Moxley G−d (2007) 1.07 (0.61, 1.90) 1.06
Chapman K−a (2006) 2.11 (1.63, 2.71) 4.36
Chapman K−b (2006) 1.25 (1.02, 1.54) 5.99
Stern AG−a (2003) 1.18 (0.60, 2.32) 0.76
Stern AG−b (2003) 1.24 (0.73, 2.12) 1.20
Stern AG−c (2003) 1.21 (0.70, 2.10) 1.12
Stern AG−d (2003) 1.21 (0.70, 2.10) 1.12
Loughlin J−a (2002) 1.26 (1.00, 1.58) 5.12
Loughlin J−b (2002) 1.34 (0.99, 1.82) 3.27
Loughlin J−c (2002) 1.22 (1.01, 1.48) 6.56
Loughlin J−d (2002) 1.13 (0.84, 1.51) 3.47
Loughlin J−e (2002) 1.49 (1.22, 1.82) 6.23
Loughlin J−f (2002) 1.15 (0.87, 1.53) 3.64
Moos V (2000) 2.25 (1.40, 3.63) 1.47
Heterogeneity test (I2 = 16.6%, P = 0.203) 1.27 (1.20, 1.35) 100.00
Z test (Z = 7.91, P < 0.001)
IL-1B
(WM+MM versus WW)
Included study OR (95CI) Weight%
Shen CK−a (2013) 1.50 (0.86, 2.61) 1.96
Shen CK−b (2013) 1.14 (0.63, 2.08) 1.72
Wu X−a (2013) 1.54 (0.75, 3.15) 1.25
Wu X−b (2013) 1.08 (0.56, 2.09) 1.47
Wu X−c (2013) 0.97 (0.51, 1.85) 1.52
Kaarvatn MH−a (2013) 1.04 (0.76, 1.41) 4.88
Kaarvatn MH−b (2013) 1.07 (0.79, 1.44) 5.06
Chen Y (2012) 0.85 (0.45, 1.61) 1.55
Jotanovic Z (2012) 1.63 (1.19, 2.24) 4.70
Solovieva S−a (2009) 1.02 (0.72, 1.45) 4.11
Solovieva S−b (2009) 2.12 (1.47, 3.05) 3.89
Solovieva S−c (2009) 1.50 (1.05, 2.13) 4.06
Ni H−a (2009) 1.19 (0.89, 1.60) 5.25
Ni H−b (2009) 1.32 (0.93, 1.87) 4.11
Ni H−c (2009) 1.39 (0.83, 2.32) 2.26
Sezgin M (2007) 1.03 (0.54, 1.94) 1.55
Moxley G−a (2007) 1.13 (0.51, 2.48) 1.05
Moxley G−b (2007) 1.48 (0.69, 3.20) 1.10
Moxley G−c (2007) 1.31 (0.62, 2.78) 1.15
Moxley G−d (2007) 1.31 (0.62, 2.78) 1.15
Chapman K−a (2006) 2.30 (1.72, 3.08) 5.26
Chapman K−b (2006) 1.25 (0.96, 1.63) 5.91
Stern AG−a (2003) 1.35 (0.61, 2.99) 1.04
Stern AG−b (2003) 0.84 (0.40, 1.74) 1.20
Stern AG−c (2003) 1.27 (0.61, 2.62) 1.22
Stern AG−d (2003) 1.27 (0.61, 2.62) 1.22
Loughlin J−a (2002) 1.22 (0.93, 1.60) 5.80
Loughlin J−b (2002) 1.36 (0.93, 1.98) 3.64
Loughlin J−c (2002) 1.25 (0.96, 1.63) 5.86
Loughlin J−d (2002) 1.10 (0.75, 1.61) 3.67
Loughlin J−e (2002) 1.58 (1.22, 2.03) 6.19
Loughlin J−f (2002) 1.15 (0.79, 1.67) 3.68
Moos V (2000) 2.14 (1.12, 4.09) 1.51
Heterogeneity test (I2 = 23.0%, P = 0.120) 1.32 (1.21, 1.43) 100.00
Z test (Z = 6.46, P < 0.001)

0.276 1 3.63 0.245 1 4.09

HWE (all P>0.05). The NOS scores of all included studies genotyping method. Subgroup analysis by disease indicated
were ≥5 and STREGA scores ≥18. The characteristics and that a higher IL-6 mutation frequency was observed among
methodological quality of the included studies are summa- knee OA and hand OA patients (knee OA—OR=2.91, 95 %
rized in Table 1. CI=1.47~5.76, P=0.002; hand OA—OR=1.63, 95 % CI=
1.04~2.55, P=0.033, respectively), but not among hip OA or
Quantitative data synthesis DIP OA patients (all P>0.05). There was also a higher IL-1A
variant frequency among hip OA patients under five genetic
Seventeen case–control studies examined genetic polymor- models (all P<0.05) and among hand OA, hip OA, and DIP
phisms of the IL-6 and IL-1 gene in OA development. Since OA patients under three genetic models (all P<0.05). A
heterogeneity was significant between the studies, the higher IL-1B polymorphism frequency was found among
random-effects model was used. The pooled estimate indicat- knee OA and hip OA patients (all P<0.05), but not among
ed that the genetic mutation frequencies of the IL-6 gene in hand OA patients (all P>0.05) (Fig. 4). Further subgroup
OA were markedly higher than those in normal controls under analyses based on genotyping method also showed that IL-6
two genetic models (allele model—OR=1.59, 95 % CI=1.21 genetic mutation frequency increased significantly with the
~2.10, P=0.001; dominant model—OR=1.91, 95 % CI= use of PCR-RFLP (allele model—OR=1.87, 95 % CI=1.22~
1.36~2.68, P<0.001) (Table 2; Fig. 3). Figure 3 also implied 2.87, P=0.004; dominant model—OR=2.03, 95 % CI=1.13
that the genetic variant frequencies of the IL-1A and IL-1B ~3.65, P=0.018, respectively) and that IL-1A and IL-1B
genes in OA were obviously higher than in the control group
under five genetic models (all P<0.05). Fig. 4 Subgroup analyses by disease and genotyping method for the„
In order to explore potential sources of heterogeneity, we relationships of IL-6, IL-1A, and IL-1B genetic polymorphisms with
also performed subgroup analyses based on disease and susceptibility to osteoarthritis under the allele model
Clin Rheumatol
IL-6
(Disease: M allele versus W allele)
Included study OR (95%CI) Weight%
Hip OA
Kolundzic R (2011) 1.23 (0.41, 3.72) 6.04
Z test (Z = 0.37, P = 0.711) 1.23 (0.41, 3.72) 6.04
Knee OA
Honsawek S (2011) 2.91 (1.47, 5.76) 15.36
Z test (Z = 3.06, P = 0.002) 2.91 (1.47, 5.76) 15.36
DIP OA
Kamarainen OP−a (2008) 1.28 (0.84, 1.96) 37.11
Kamarainen OP−b (2008) 1.55 (0.59, 4.05) 7.90
Heterogeneity test (I2 = 0.00%, P = 0.723) 1.32 (0.89, 1.95) 45.01
Z test (Z = 1.40, P = 0.163)
Hand OA
Pola E (2005 1.63 (1.04, 2.55) 33.59
Z test (Z = 2.13, P = 0.033) 1.63 (1.04, 2.55) 33.59
Heterogeneity test (I2 = 5.3%, P = 0.377) 1.59 (1.21, 2.10) 100.00
Z test (Z = 3.33, P = 0.001)
0.174 1 5.76
IL-1A
(Disease: M allele versus W allele)
Included study OR (95%CI) Weight%
Knee OA
Kaarvatn MH−a (2013) 1.22 (0.97, 1.54) 19.16
Sezgin M (2007) 1.60 (0.95, 2.70) 4.02
Loughlin J−b (2002) 1.38 (1.04, 1.82) 13.41
Heterogeneity test (I2 = 0.00%, P = 0.598) 1.31 (1.11, 1.56) 36.60
Z test (Z = 3.15, P = 0.002)
Hip OA
Kaarvatn MH−b (2013) 1.02 (0.81, 1.29) 19.61
Loughlin J−a (2002) 1.01 (0.83, 1.24) 24.99
2
Heterogeneity test (I = 0.00%, P = 0.944) 1.02 (0.88, 1.19) 44.60
Z test (Z = 0.24, P = 0.812)
DIP OA
Solovieva S (2009) 1.04 (0.80, 1.34) 15.59
Z test (Z = 0.27, P = 0.789) 1.04 (0.80, 1.34) 15.59
Hand OA
Stern AG (2003) 1.05 (0.59, 1.89) 3.21
Loughlin J−b (2002) 1.05 (0.59, 1.89) 3.21
Z test (Z = 0.17, P = 0.865)
Heterogeneity test (I2 = 5.2%, P = 0.388) 1.12 (1.01, 1.25) 100.00
Z test (Z = 2.15, P = 0.032)
0.37 1 2.7
IL-1B
(Disease: M allele versus W allele)
Included study OR (95%CI) Weight%
Knee OA
Shen CK−a (2013) 1.05 (0.74, 1.50) 2.53
Shen CK−b (2013) 1.11 (0.78, 1.58) 2.54
Wu X−a (2013) 1.48 (0.80, 2.72) 0.92
Wu X−b (2013) 1.14 (0.71, 1.80) 1.55
Wu X−c (2013) 1.06 (0.65, 1.73) 1.42
Kaarvatn MH−a (2013) 1.07 (0.83, 1.38) 4.34
Chen Y (2012) 1.22 (0.72, 2.06) 1.22
Jotanovic Z (2012) 1.50 (1.18, 1.90) 4.87
Ni H−a (2009) 1.19 (0.99, 1.43) 7.02
Ni H−b (2009) 1.29 (1.03, 1.60) 5.39
Ni H−c (2009) 1.36 (0.99, 1.88) 2.94
Sezgin M (2007) 1.04 (0.61, 1.76) 1.21
Loughlin J−b (2002) 1.34 (0.99, 1.82) 3.27
Loughlin J−d (2002) 1.13 (0.84, 1.51) 3.47
Loughlin J−f (2002) 1.15 (0.87, 1.53) 3.64
Moos V (2000) 2.25 (1.40, 3.63) 1.47
Heterogeneity test (I2 = 0.00%, P = 0.534) 1.24 (1.14, 1.34) 47.80
Z test (Z = 5.40, P < 0.001)
Hip OA
Kaarvatn MH−b (2013) 1.09 (0.85, 1.39) 4.54
Chapman K−a (2006) 2.11 (1.63, 2.71) 4.36
Chapman K−b (2006) 1.25 (1.02, 1.54) 5.99
Loughlin J−a (2002) 1.26 (1.00, 1.58) 5.12
Loughlin J−c (2002) 1.22 (1.01, 1.48) 6.56
Loughlin J−e (2002) 1.49 (1.22, 1.82) 6.23
Heterogeneity test (I2 = 71.6%, P = 0.003) 1.36 (1.15,) 1.61 32.81
Z test (Z = 3.63, P < 0.001)
DIP OA
Solovieva S−a (2009) 1.09 (0.82, 1.45) 3.60
Solovieva S−b (2009) 1.61 (1.19, 2.18) 3.26
Solovieva S−c (2009) 1.22 (0.95, 1.58) 4.30
Heterogeneity test (I2 = 43.9%, P = 0.168) 1.28 (1.03, 1.59) 11.16
Z test (Z = 2.23, P = 0.026)
Hand OA
Moxley G−a (2007) 1.10 (0.56, 2.14) 0.78
Moxley G−b (2007) 1.35 (0.78, 2.34) 1.13
Moxley G−c (2007) 1.07 (0.61, 1.90) 1.06
Moxley G−d (2007) 1.07 (0.61, 1.90) 1.06
Stern AG−a (2003) 1.18 (0.60, 2.32) 0.76
Stern AG−b (2003) 1.24 (0.73, 2.12) 1.20
Stern AG−c (2003) 1.21 (0.70, 2.10) 1.12
Stern AG−d (2003) 1.21 (0.70, 2.10) 1.12
Heterogeneity test (I2 = 0.00%, P = 0.999) 1.18 (0.96, 1.45) 8.23
Z test (Z = 1.62, P = 0.106)
Heterogeneity test (I2 = 16.6%, P = 0.203) 1.27 (1.20, 1.35) 100.00
Z test (Z = 7.91, P < 0.001)
0.276 1 3.63
IL-6
(Genotyping method: M allele versus W allele)
Included study OR (95%CI) Weight%
PCR−RFLP
Kolundzic R (2011) 1.23 (0.41, 3.72) 6.04
Honsawek S (2011) 2.91 (1.47, 5.76) 15.36
Pola E (2005) 1.63 (1.04, 2.55) 33.59
Heterogeneity test (I2 = 20.3%, P = 0.285) 1.87 (1.22, 2.87) 54.99
Z test (Z = 2.89, P = 0.004)
Non−PCR−RFLP
Kamarainen OP−a (2008) 1.28 (0.84, 1.96) 37.11
Kamarainen OP−b (2008) 1.55 (0.59, 4.05 7.90
Subtotal (I2 = 0.00%, P = 0.723) 1.32 (0.89, 1.95 45.01
Z test (Z = 1.40, P = 0.163)
Heterogeneity test (I2 = 5.3%, P = 0.377) 1.59 (1.21, 2.10) 100.00
Z test (Z = 3.33, P = 0.001)
0.174 1 5.76
IL-1A
(Genotyping method: M allele versus W allele)
Included study OR (95%CI) Weight%
Non−PCR−RFLP
Kaarvatn MH−a (2013) 1.22 (0.97, 1.54) 19.16
Kaarvatn MH−b (2013) 1.02 (0.81, 1.29) 19.61
Stern AG (2003) 1.05 (0.59, 1.89) 3.21
Heterogeneity test (I2 = 0.00%, P = 0.561) 1.11 (0.95, 1.30) 41.98
Z test (Z = 1.34, P = 0.181)
PCR−RFLP
Solovieva S (2009) 1.04 (0.80, 1.34) 15.59
Sezgin M (2007) 1.60 (0.95, 2.70) 4.02
Loughlin J−a (2002) 1.01 (0.83, 1.24) 24.99
Loughlin J−b (2002) 1.38 (1.04, 1.82) 13.41
Heterogeneity test (I2 = 41.8%, P = 0.161) 1.16 (0.96, 1.40) 58.02
Z test (Z = 1.54, P = 0.124)
Heterogeneity test (I2 = 5.2%, P = 0.388) 1.12 (1.01, 1.25) 100.00
Z test (Z = 2.15, P = 0.032)
0.37 1 2.7
IL-1B
(Genotyping method: M allele versus W allele)
Included study OR (95%CI) Weight%
PCR−RFLP
Shen CK−a (2013) 1.05 (0.74, 1.50) 2.53
Shen CK−b (2013) 1.11 (0.78, 1.58) 2.54
Chen Y (2012) 1.22 (0.72, 2.06) 1.22
Solovieva S−a (2009) 1.09 (0.82, 1.45) 3.60
Solovieva S−b (2009) 1.61 (1.19, 2.18) 3.26
Solovieva S−c (2009) 1.22 (0.95, 1.58) 4.30
Ni H−a (2009) 1.19 (0.99, 1.43) 7.02
Ni H−b (2009) 1.29 (1.03, 1.60) 5.39
Ni H−c (2009) 1.36 (0.99, 1.88) 2.94
Sezgin M (2007) 1.04 (0.61, 1.76) 1.21
Chapman K−a (2006) 2.11 (1.63, 2.71) 4.36
Chapman K−b (2006) 1.25 (1.02, 1.54) 5.99
Loughlin J−a (2002) 1.26 (1.00, 1.58) 5.12
Loughlin J−b (2002) 1.34 (0.99, 1.82) 3.27
Loughlin J−c (2002) 1.22 (1.01, 1.48) 6.56
Loughlin J−d (2002) 1.13 (0.84, 1.51) 3.47
Loughlin J−e (2002) 1.49 (1.22, 1.82) 6.23
Loughlin J−f (2002) 1.15 (0.87, 1.53) 3.64
Moos V (2000) 2.25 (1.40, 3.63) 1.47
Heterogeneity test (I2 = 41.2%, P = 0.032) 1.30 (1.20, 1.41) 74.12
Z test (Z = 6.31, P < 0.001)
Non−PCR−RFLP
Wu X−a (2013) 1.48 (0.80, 2.72) 0.92
Wu X−b (2013) 1.14 (0.71, 1.80) 1.55
Wu X−c (2013) 1.06 (0.65, 1.73) 1.42
Kaarvatn MH−a (2013) 1.07 (0.83, 1.38) 4.34
Kaarvatn MH−b (2013) 1.09 (0.85, 1.39) 4.54
Jotanovic Z (2012) 1.50 (1.18, 1.90) 4.87
Moxley G−a (2007) 1.10 (0.56, 2.14) 0.78
Moxley G−b (2007) 1.35 (0.78, 2.34) 1.13
Moxley G−c (2007) 1.07 (0.61, 1.90) 1.06
Moxley G−d (2007) 1.07 (0.61, 1.90) 1.06
Stern AG−a (2003) 1.18 (0.60, 2.32) 0.76
Stern AG−b (2003) 1.24 (0.73, 2.12) 1.20
Stern AG−c (2003) 1.21 (0.70, 2.10) 1.12
Stern AG−d (2003) 1.21 (0.70, 2.10) 1.12
Heterogeneity test (I2 = 0.00%, P = 0.942) 1.20 (1.08, 1.34) 25.88
Z test (Z = 3.31, P = 0.001)
Heterogeneity test (I2 = 16.6%, P = 0.203) 1.27 (1.20, 1.35) 100.00
Z test (Z = 7.91, P < 0.001)
0.276 1 3.63
 Clin Rheumatol

Fig. 5 Sensitivity analysis of the IL-6

summary odds ratio coefficients (M allele versus W allele)
Lower CI Limi t
on the relationships of IL-6, IL- Estimate Upper CI Limit

1A, and IL-1B genetic Kolundzic R (2011)

polymorphisms with
susceptibility to osteoarthritis
under the allele model
Honsawek S (2011)

Kamarainen OP−a (2008)

Kamarainen OP−b (2008)

Pola E (2005)

1.07 1.21 1.59 2.10 2.53

IL-1A
(M allele versus W allele)
Lower CI Limit Estimate Upper CI Limit

Kaarvatn MH−a (2013)

Kaarvatn MH−b (2013)

Solovieva S (2009)

Sezgin M (2007)

Stern AG (2003)

Loughlin J−a (2002)

Loughlin J−b (2002)

0.97 1.01 1.12 1.25 1.31

IL-1B
(M allele versus W allele)
Lower CI Limit Estimate Upper CI Limit

Shen CK−a (2013)

Shen CK−b (2013)
Wu X−a (2013)
Wu X−b (2013)
Wu X−c (2013)
Kaarvatn MH−a (2013)
Kaarvatn MH−b (2013)
Chen Y (2012)
Jotanovic Z (2012)
Solovieva S−a (2009)
Solovieva S−b (2009)
Solovieva S−c (2009)
Ni H−a (2009)
Ni H−b (2009)
Ni H−c (2009)
Sezgin M (2007)
Moxley G−a (2007)
Moxley G−b (2007)
Moxley G−c (2007)
09 Moxley G−d (2007)
Chapman K−a (2006)
Chapman K−b (2006)
Stern AG−a (2003)
Stern AG−b (2003)
Stern AG−c (2003)
Stern AG−d (2003)
Loughlin J−a (2002)
Loughlin J−b (2002)
Loughlin J−c (2002)
Loughlin J−d (2002)
Loughlin J−e (2002)
Loughlin J−f (2002)
Moos V (2000)

1.18 1.20 1.27 1.35 1.37

Clin Rheumatol

IL-6
variant frequency was enhanced markedly in both PCR-RFLP (M allele versus W allele)
2
and non-PCR-RFLP groups (Fig. 4). (Egger’s test: t = 0.40, P = 0.718)

Evaluation of heterogeneity and publication bias

1

Sensitivity analysis was performed to assess the influence of

Log[OR]
each individual study on the pooled ORs by omitting each
individual study in turn. The analysis results suggest that no 0

individual studies obviously influenced the pooled ORs

(Fig. 5). Funnel plots and Egger’s linear regression test were
used to assess potential publication bias of included studies.
−1
The shapes of the funnel plots did not reveal any evidence of 0 0.2 0.4 0.6

obvious asymmetry (Fig. 6). Egger’s test also did not show SE Log[OR]

strong statistical evidence of publication bias (all P>0.05).

IL-1A
(M allele versus W allele)
1 (Egger’s test: t = 1.17, P = 0.295)

Discussion

In the present meta-analysis, we aimed to clarify the relation- 0.5

ships between genetic polymorphisms in the IL-1A, IL-1B,

and IL-6 genes and the development of OA. The findings of Log[OR]

our meta-analysis illustrate that genetic variants in the IL-6 0

gene was positively correlated with the pathogenesis of OA,
indicating that carrying IL-6 genetic polymorphisms might be
a dominant risk factor for the development of OA. It is widely
−0.5
accepted that IL-6 is primarily produced at sites of acute and
0 0.1 0.2 0.3
chronic inflammation, where it is secreted into the serum and SE Log[OR]

induces a transcriptional inflammatory response through in-

terleukin 6 receptor alpha [17]. Furthermore, inflammatory IL-1A
(M allele versus W allele)
mechanisms play a crucial role in the etiology and progress 1
(Egger’s test: t = -0.58, P = 0.565)
of cartilage degradation, which is the main characteristic of
OA [20]. IL-6 is capable of increasing the number of inflam-
matory cells in the synovial tissue, promoting proliferation of 0.5

chondrocytes, leading to the amplification of IL-1, and there-

Log[OR]

by resulting in elevated production of matrix metalloprotein-

ases and inhibiting the production of proteoglycans. Genetic
variations in the IL-6 gene, especially within the non-coding
0

promoter sequence, may have an important influence on the

expression of IL-6 [22]. Therefore, it is plausible that IL-6
genetic polymorphisms might be implicated in the pathogen- −0.5

esis of OA. Kamarainen et al. concluded that the frequency of
G alleles of two promoter single nucleotide polymorphisms
(SNPs) G-597A and G-174C in the IL-6 gene were higher in
OA patients than in healthy controls, suggesting that the
presence of G alleles at common IL-6 polymorphic promoter
loci was closely connected with more severe OA outcomes
[22].
Furthermore, our meta-analysis results demonstrate that
genetic polymorphisms in the IL-1A and IL-1B genes might
be closely linked to the occurrence of OA, revealing that IL-
1A, and IL-1B, genetic polymorphisms might significantly
contribute to an increased risk of OA. Nevertheless, the exact
mechanism by which IL-1A and IL-1B genetic polymorphisms
0 0.2 0.4
SE Log[OR]
Fig. 6 Funnel plot of publication biases on the relationships of IL-6, IL-
1A, and IL-1B genetic polymorphisms with susceptibility to osteoarthritis
under the allele model
induce the risk of OA has remained largely unknown. A
reasonable explanation may be that common allelic variants
in IL-1A and IL-1B gene may have a crucial impact on the
transcription and synthesis of IL-1A and IL-1B, which might
result in cartilage breakdown and promote the release of
arachidonic acid from cell membranes and the subsequent
production of prostanoids and eicosanoids, and consequently
give rise to the progress of OA [32]. A previous study

conducted by Stern et al. also showed significant correlation
between erosive hand OA and a genomic region containing
the IL-1B 5810 SNP in a US Caucasian population, supporting
a potential role for IL-1 in the etiology of a severe phenotype
of hand OA [14].
Stratified analysis based on the type of disease was care-
fully performed when the possibility of existing apparent
heterogeneity was taken into consideration. The results of
subgroup analysis by the type of disease illustrated that the
presence of IL-6 genetic polymorphisms was positively con-
nected with the pathogenesis of knee OA and hand OA, but
not with hip or DIP OA. In addition, our findings also revealed
close relationships between IL-1B genetic polymorphisms and
the development of knee, hip, and DIP OA and a higher IL-1A
variant frequency among hip OA, hand OA, hip OA, and DIP
OA patients. In short, our meta-analysis results were in line
with previous studies in finding that genetic polymorphisms in
IL-1A, IL-1B, and IL-6 genes might be strongly correlated
with the pathogenesis of OA, revealing that these polymor-
phisms may be valuable biomarkers for the early detection of
OA.
Although the relatively large sample size and confor-
mation to HWE strengthen our meta-analysis, there are
some limitations that should be kept in mind. Firstly,
the potential confounding effect of sex, disease, and
ethnicity was uncontrolled in a great portion of the
included studies, which limits the explanatory power
of our findings and the subsequent subgroup analyses.
Secondly, due to our search methods and our inclusion
and exclusion criteria, it is likely that there were un-
published studies and articles in languages other than
English and Chinese that were not included. In this
regard, although the Egger’s test and funnel plots failed
to reveal any evidence of publication bias in our meta-
analysis, the possibility of bias cannot be ruled out
completely. A third limitation of our meta-analysis is
the crude division criteria of ethnic groups into
“Caucasian,” “Asian,” or “African” and the fact that
our analyses only focused on the Asian and Caucasian
populations. Further studies from different populations
are warranted to clarify the present results. Another
potential limitation is that our meta-analysis may still
be underpowered. Also, our inability to acquire original
data from the included studies limited the depth and
detail of our meta-analysis. Despite these limitations,
this is the first meta-analysis on associations between
IL-6, IL-1A, and IL-1B genetic polymorphisms and OA.
More importantly, our meta-analysis was specific to
three polymorphisms in the IL family and accounts for
possible synergistic effects of other candidate genes or
polymorphisms. Besides, we rigorously quantified and
analyzed previous studies’ inconsistent results, providing
a more reliable conclusion.
Clin Rheumatol
Taken together, our findings provide evidence that IL-6, IL-
1A, and IL-1B genetic polymorphisms may be correlated with
susceptibility to OA. Thus, IL-6, IL-1A, and IL-1B genetic
polymorphisms may be useful for identifying OA patients at
an early stage. For now, more studies among different ethnic-
ities are required to validate these findings and strengthen our
conclusions.
Acknowledgments We would like to acknowledge the reviewers for
their helpful comments on this paper.
Competing interests The authors have declared that no competing
interests exist.
References
1. Ni H, Shi D, Dai J, Qin J, Xu Y, Zhu L et al (2009) Genetic
polymorphisms of interleukin-1beta (-511C/T) and interleukin-1 re-
ceptor antagonist (86-bpVNTR) in susceptibility to knee osteoarthri-
tis in a Chinese Han population. Rheumatol Int 29(11):1301–1305
2. Attur M, Wang HY, Kraus VB, Bukowski JF, Aziz N, Krasnokutsky
S et al (2010) Radiographic severity of knee osteoarthritis is condi-
tional on interleukin 1 receptor antagonist gene variations. Ann
Rheum Dis 69(5):856–861
3. Fransen M, Bridgett L, March L, Hoy D, Penserga E, Brooks P
(2011) The epidemiology of osteoarthritis in Asia. Int J Rheum Dis
14(2):113–121
4. Pereira D, Peleteiro B, Araujo J, Branco J, Santos RA, Ramos E
(2011) The effect of osteoarthritis definition on prevalence and inci-
dence estimates: a systematic review. Osteoarthr Cartil 19(11):1270–
1285
5. Kaarvatn MH, Jotanovic Z, Mihelic R, Etokebe GE, Mulac-Jericevic
B, Tijanic T et al (2013) Associations of the interleukin-1 gene locus
polymorphisms with risk to hip and knee osteoarthritis: gender and
subpopulation differences. Scand J Immunol 77(2):151–161
6. Jotanovic Z, Etokebe GE, Mihelic R, Kaarvatn MH, Mulac-Jericevic
B, Tijanic T et al (2012) IL1B–511(G > A) and IL1RN (VNTR)
allelic polymorphisms and susceptibility to knee osteoarthritis in
Croatian population. Rheumatol Int 32(7):2135–2141
7. Lv C, Xu X, Wang J, Zhang Z, Zhang D, Guo C et al (2012)
Combined effect of cytokine gene polymorphisms on end-stage knee
osteoarthritis from Chinese Han population. Rheumatol Int 32(11):
3625–3629
8. Honsawek S, Deepaisarnsakul B, Tanavalee A, Yuktanandana P,
Bumrungpanichthaworn P, Malila S et al (2011) Association of the
IL-6–174G/C gene polymorphism with knee osteoarthritis in a Thai
population. Genet Mol Res 10(3):1674–1680
9. Brocker C, Thompson D, Matsumoto A, Nebert DW, Vasiliou V
(2010) Evolutionary divergence and functions of the human interleu-
kin (IL) gene family. Hum Genomics 5(1):30–55
10. Akdis M, Burgler S, Crameri R, Eiwegger T, Fujita H, Gomez E et al
(2011) Interleukins, from 1 to 37, and interferon-gamma: receptors,
functions, and roles in diseases. J Allergy Clin Immunol 127(3):701–
721, e1-70
11. Sims JE, Smith DE (2010) The IL-1 family: regulators of immunity.
Nat Rev Immunol 10(2):89–102
12. Moxley G, Han J, Stern AG, Riley BP (2007) Potential
influence of IL1B haplotype and IL1A-IL1B-IL1RN extended
haplotype on hand osteoarthritis risk. Osteoarthr Cartil 15(10):
1106–1112
Clin Rheumatol
13. Benito MJ, Veale DJ, FitzGerald O, van den Berg WB, Bresnihan B
(2005) Synovial tissue inflammation in early and late osteoarthritis.
Ann Rheum Dis 64(9):1263–1267
14. Stern AG, de Carvalho MR, Buck GA, Adler RA, Rao TP, Disler D
et al (2003) Association of erosive hand osteoarthritis with a single
nucleotide polymorphism on the gene encoding interleukin-1 beta.
Osteoarthr Cartil 11(6):394–402
15. Lafage M, Maroc N, Dubreuil P, de Waal Malefijt R, Pebusque MJ,
Carcassonne Y et al (1989) The human interleukin-1 alpha gene is
located on the long arm of chromosome 2 at band q13. Blood 73(1):
104–107
16. Sezgin M, Erdal ME, Altintas ZM, Ankarali HC, Barlas IO, Turkmen
E et al (2007) Lack of association polymorphisms of the IL1RN,
IL1A, and IL1B genes with knee osteoarthritis in Turkish patients.
Clin Invest Med 30(2):E86–E92
17. Gabay C (2006) Interleukin-6 and chronic inflammation. Arthritis
Res Ther 8(Suppl 2):S3
18. Doss F, Menard J, Hauschild M, Kreutzer HJ, Mittlmeier T, Muller-
Steinhardt M et al (2007) Elevated IL-6 levels in the synovial fluid of
osteoarthritis patients stem from plasma cells. Scand J Rheumatol
36(2):136–139
19. Kolundzic R, Trkulja V, Mikolaucic M, Kolundzic MJ, Pavelic SK,
Pavelic K (2011) Association of interleukin-6 and transforming
growth factor-beta1 gene polymorphisms with developmental hip
dysplasia and severe adult hip osteoarthritis: a preliminary study.
Cytokine 54(2):125–128
20. Pola E, Papaleo P, Pola R, Gaetani E, Tamburelli FC, Aulisa
L et al (2005) Interleukin-6 gene polymorphism and risk of
osteoarthritis of the hip: a case-control study. Osteoarthr Cartil
13(11):1025–1028
21. Biffl WL, Moore EE, Moore FA, Peterson VM (1996) Interleukin-6
in the injured patient. Marker of injury or mediator of inflammation?
Ann Surg 224(5):647–664
22. Kamarainen OP, Solovieva S, Vehmas T, Luoma K,
Riihimaki H, Ala-Kokko L et al (2008) Common
interleukin-6 promoter variants associate with the more se-
vere forms of distal interphalangeal osteoarthritis. Arthritis
Res Ther 10(1):R21
23. Little J, Higgins JP, Ioannidis JP, Moher D, Gagnon F, von Elm E et al
(2009) Strengthening the Reporting of Genetic Association studies
(STREGA)—an extension of the STROBE statement. Eur J Clin
Investig 39(4):247–266
24. Stang A (2010) Critical evaluation of the Newcastle-Ottawa scale for
the assessment of the quality of nonrandomized studies in meta-
analyses. Eur J Epidemiol 25(9):603–605
25. Jackson D, White IR, Riley RD (2012) Quantifying the impact of
between-study heterogeneity in multivariate meta-analyses. Stat Med
31(29):3805–3820
26. Peters JL, Sutton AJ, Jones DR, Abrams KR, Rushton L (2006)
Comparison of two methods to detect publication bias in meta-
analysis. JAMA 295(6):676–680
27. Zintzaras E, Ioannidis JP (2005) HEGESMA: genome search
meta-analysis and heterogeneity testing. Bioinformatics 21(18):
3672–3673
28. Chapman K, Loughlin J (2006) Association of the interleukin-1 gene
cluster with osteoarthritis of the hip: comment on the article by
Meulenbelt et al. and the letter by Smith et al. Arthritis Rheum
54(11):3722–3723
29. Chen Y (2012) Association of interleukin-1β gene polymorphisms
and susceptibility to primary osteoarthritis of the knee. Chin J
Gerontol 32(16):3425–3426
30. Loughlin J, Dowling B, Mustafa Z, Chapman K (2002) Association
of the interleukin-1 gene cluster on chromosome 2q13 with knee
osteoarthritis. Arthritis Rheum 46(6):1519–1527
31. Shen CK, Lv CY, Wang YZ, Zhang HN, Meng F (2013) Association
of interleukin-1β gene polymorphisms and susceptibility to advanced
knee osteoarthritis in Han population. Chin J Joint Surg Electron
Version 7(3):378–385
32. Solovieva S, Kamarainen OP, Hirvonen A, Hamalainen S, Laitala M,
Vehmas T et al (2009) Association between interleukin 1 gene cluster
polymorphisms and bilateral distal interphalangeal osteoarthritis. J
Rheumatol 36(9):1977–1986
33. Wu X, Kondragunta V, Kornman KS, Wang HY, Duff GW, Renner
JB et al (2013) IL-1 receptor antagonist gene as a predictive biomark-
er of progression of knee osteoarthritis in a population cohort.
Osteoarthr Cartil 21(7):930–938
34. Moos V, Rudwaleit M, Herzog V, Hohlig K, Sieper J, Muller B
(2000) Association of genotypes affecting the expression of
interleukin-1beta or interleukin-1 receptor antagonist with osteoar-
thritis. Arthritis Rheum 43(11):2417–2422