Clinical Rheumatology

https://doi.org/10.1007/s10067-020-05150-z

ORIGINAL ARTICLE

Serum levels of leptin, osteopontin, and sclerostin in patients

with and without knee osteoarthritis
Sicong Min 1,2,3 & Tianshu Shi 1,2 & Xiao Han 1,2 & Dongyang Chen 1,2 & Zhihong Xu 1,2 & Dongquan Shi 1,2 & Huajian Teng 1,2 &
Qing Jiang 1,2

Received: 26 November 2019 / Revised: 13 April 2020 / Accepted: 5 May 2020

# International League of Associations for Rheumatology (ILAR) 2020

Abstract
Objective To investigate the relationship between leptin, osteopontin (OPN), sclerostin (SOST) and severity of knee osteoar-
thritis (KOA).
Methods The study included 148 consecutive patients with knee OA and 101 non-KOA subjects enrolled in this cross-sectional
study. All patients fulfilled the American College of Rheumatology criteria for primary knee OA. Severity of the disease was
assessed using plain radiography of the affected knee, according to the Kellgren and Lawrence classification. Fasting blood
samples were obtained from all patients and controls; the serum samples were kept at − 80 °C before assessment of leptin, OPN,
and SOST using a multiplex particle-based flow cytometric assay.
Results KOA patients group compared with the control group, serum leptin (KOA, 26581.7 ± 2011.5 pg/ml, vs control,6936.4 ±
702.2 pg/ml),OPN (KOA, 4908.3 ± 769.4 pg/ml, vs control, 2182.5 ± 217.8 pg/ml), and SOST (KOA, 2481.9 ± 543.5 pg/ml, vs
control, 1288.9 ± 267.7 pg/ml) in the KOA group were higher than control group; there were also differences in three bone
metabolic factors between male and female in the KOA group; meanwhile, there was correlation between each factor and the
incidence of KOA.
Conclusion Our study of 249 serum samples was conducted. Serum leptin, OPN, and SOST were significantly increased in KOA
patients, and there was an internal correlation; these findings could, at best, contribute to the identification of novel targets for
medical interventions.

Key Points
• The aim of this study was to assess the relationships of radiographic knee OA with altered serum levels of leptin, OPN, and SOST. Our study of 249
serum samples was conducted. Serum leptin, OPN, and SOST were significantly increased in KOA patients compared with control group. There were
gender differences in the concentration of three serum bone turnover factors in KOA group and control group. Serum SOST concentration increased
with Kellgren-Lawrence (K-L) grading. We found that serum leptin, OPN, and SOST were significantly increased in KOA patients, and there was an
internal correlation. Leptin had a remarkable diagnostic value in the incidence of KOA.

Keywords Serum . Leptin . Knee osteoarthritis . Osteopontin . Sclerostin

* Huajian Teng Dongyang Chen

[email protected] [email protected]
* Qing Jiang
[email protected] Zhihong Xu
Sicong Min [email protected]
[email protected]

Tianshu Shi Dongquan Shi

[email protected] [email protected]

Xiao Han Extended author information available on the last page of the article
[email protected]
 Clin Rheumatol

Abbreviations macrophages, activated T cells, chondrocytes, and osteoblasts

OA Osteoarthritis [8]. OPN may be involved in the molecular pathogenesis of
KOA knee osteoarthritis OA, causing progressive degeneration of articular cartilage
OPN osteopontin [9]. Expression level of OPN mRNA isolated from human
SOST sclerostin: OA cartilage and chondrocytes was increased compared with
K-L Kellgren-Lawrence normal cartilage [10]. Leptin is a 16-kDa (kilodalton) protein
BMI body mass index commonly found in different tissues, which is primarily se-
creted by placenta and white adipose tissue [11]. Leptin con-
trols the amount of adipose tissue and BMI, adjusts food in-
Introduction take, and stimulates energy consumption at hypothalamic nu-
clei [12]. Additionally, it promotes collagen production and
Osteoarthritis (OA) is a chronic degenerative joint disease, ossification in bone and reproduction of osteoblast [13].
resulting in physical disability, substantial morbidity, and de- Leptin may participate in the inflammatory response, which
creased quality of life. OA is the most common kind of arthri- further shows that leptin may play a substantial role in the
tis mainly characterized by focal lesions of the articular carti- development of OA [14].Serum levels of bone turnover bio-
lage, combined with a hypertrophic reaction (sclerosis) in the markers have been widely investigated in patients with knee
subchondral bone and new bone formation (osteophytes) at OA, in search for biochemical markers that could determine
the joint margins and mild, chronic nonspecific inflammatory disease severity and progression. However, there have been
synovitis [1]. several conflicting studies on the possible association between
To identify patients with a high risk and develop new treat- serum levels of SOST, OPN, and leptin with disease activity
ments for destructive OA, further sensitive techniques, in in primary knee OA. Meanwhile, the relationship between
comparison with plain X-rays, are required. To date, the gold these biomarkers still remains elusive. To date, animal-based
standard is still the Kellgren-Lawrence (K-L) grading scale, studies have been frequently conducted, while a limited num-
which is based on the measurement of joint space width on ber of studies have analyzed comprehensively those bio-
standard plain X-ray [2]. This parameter could not provide markers in human serum. Thus, the aim of the present study
enough sensitivity and reproducibility, and that fails to detect was to evaluate these markers and their relationship with each
the early metabolic changes preceding the appearance of other, as well as their relationship with the severity of OA.
structural changes. Specific and sensitive biochemical
markers reflecting abnormalities of the turnover of bone, car-
tilage, and synovial tissues may be helpful for investigating Methods
and monitoring of OA [3]. Meanwhile, there are academic and
industrial demands for new tools to detect the early metabolic KOA patients and non-KOA subjects
lesions occurring in joint tissues before the appearance of
changes in the medical images. Development of these tools The study included 148 consecutive patients with knee OA
may assist healthcare providers to manage earlier treatment enrolled in this cross-sectional study (M/F, 36/112; age, 68.0
and further improve therapeutic efficacy. Healthcare providers (43–87) years) who were attending the Department of Sports
also need further sensitive approaches to evaluate response to Medicine and Adult Reconstructive Surgery, Drum Tower
treatment. To fit these purposes, soluble biomarkers may be Hospital (Nanjing, China) for undergoing a total knee
appropriate candidates. With development of clinical bioinfor- arthroplasty surgery. All patients fulfilled the American
matics for medical inspection, the scientific community’s in- College of Rheumatology criteria for primary knee OA [15].
terest in OA biochemical markers has increased during the last All KOA patients were excluded from a history of heart dis-
decade. ease, hypertension, diabetes, chronic kidney disease, autoim-
There are several biomarkers most commonly presented in mune disease, and fracture. None of the patients was receiving
recent studies, including leptin, osteopontin (OPN), and treatment that might interfere with bone metabolism, such as
sclerostin (SOST). SOST is encoded by SOST gene and is estrogen replacement therapy, thyroid replacement therapy,
specifically expressed in osteocyte [4]. It blocks the prolifer- diuretics or other drugs that may influence the serum bio-
ation and early differentiation of osteoblast through its nega- markers mentioned above.
tive effect on Wnt signaling pathway [5]. Wafa et al. [6] con- Non-KOA subjects included 101 women and men with a
firmed that in cartilage of mice, SOST maintains the integrity mean (range) age of 57.7(41–79) years. Women and men were
by preserving its metabolism. Lower level of SOST in plasma randomly selected from outpatient who were taking a medical
and synovial fluid has been reported to be related to less se- examination. None of the subjects had evidence of symptom-
verity of OA [7]. OPN is one of the major noncollagenous atic and radiography OA as assessed by clinical examination
bone matrix proteins generated in various cell types, including of the knees performed by an experienced orthopedist and by
Clin Rheumatol

the answer to the following question: “Has a doctor ever told distributed continuous variables, and Mann-Whitney U test
you that you had osteoarthritis?”. All subjects were healthy when the distribution was skewed. p < 0.05 was considered
without any disease (heart disease, hypertension, diabetes, statistically significant. The power analysis was statistically
chronic, kidney disease, autoimmune disease, and fracture) analyzed by using G*Power 3.1.9.2 software (Dusseldorf). G
or treatment that might interfere with bone or joint metabo- power post hoc analysis showed t test power(1-β err
lism, including hormone replacement therapy in postmeno- prob):0.9999875 and X2 test power(1-β err prob):0.9972310.
pausal women. Effects of different variables on knee osteoarthritis (KOA)
were assessed by univariate regression analysis. Variables,
Kellgren and Lawrence classification in which the unadjusted p < 0.05 in logistic regression analy-
sis, were identified as potential risk markers and included in
Severity of the disease was assessed using plain radiography the full model. We reduced the model using multivariate lo-
of the affected knee, according to the Kellgren and Lawrence gistic regression analysis and eliminated potential markers
classification. The K-L grading system was used for classify- using likelihood-ratio test. An exploratory evaluation of addi-
ing radiographic osteoarthritis. It uses four radiographic fea- tional cut points was performed using receiver operating char-
tures: joint space narrowing, osteophytes, subchondral sclero- acteristics (ROC) curve analysis.
sis, and subchondral cysts. The severity of radiographic
changes increases from grade 0 to 4 with grade 0 meaning
no radiographic features of osteoarthritis, whereas grade 4 Results
means large osteophytes, marked joint space narrowing, se-
vere sclerosis, and definite bony deformity [16]. Baseline clinical parameters

Biochemical measurements The patients’ demographic and clinical characteristics are pre-
sented in Table 1. One hundred forty-eight KOA patients (122
Fasting blood samples were obtained from all patients with women, 26 men; mean (range) age 68.0 (43–87) years) and
OA within 1 month, and radiographs were acquired. The same 101 non-KOA controls (51 women and 38 men with a mean
types of sample were obtained from controls. The serum was (range) age of 57.7 (41–79) years) were included in this study.
thawed and aliquoted into 1.5-ml volumes and then stored at It was revealed that the control group involved significantly
− 80 °C until being analyzed. Markers of bone and cartilage further female patients than male patients compared with
turnover serum samples were kept at − 80 °C before assess- KOA patients, and the patients’ age and weight in the KOA
ment of leptin, OPN, and SOST using a multiplex particle- group were notably higher than those in the control group. The
based flow cytometric assay (Millipore Corporation, Billerica, KOA patients’ height was lower than that in the control group,
MA 01821, USA). This method is well correlated with stan- and body mass index (BMI) was higher in the KOA group
dard procedures (ELISA). Standard curves of known concen- than that in the control group, while there was no significant
trations of recombinant human cytokines/chemokines were difference(P > 0.05).Serum levels of leptin (KOA, 26581.7 ±
used to convert fluorescence units to cytokine concentration 2011.5 pg/ml vs control, 6936.4 ± 702.2 pg/ml), OPN (KOA,
units (pg/ml). Minimum detectable concentrations, based on 4908.3 ± 769.4 pg/ml vs control, 2182.5 ± 217.8 pg/ml), and
manufacturer specifications Leptin = 30.28 pg/ml, OPN = SOST (KOA, 2481.9 ± 543.5 pg/ml vs control, 1288.9 ±
92.11pg/ml and SOST = 24.08pg/ml. A good agreement be- 267.7 pg/ml) in the KOA patients group were significant
tween signal and cytokine was found within the assay range higher than those in the control group.
(R 2 ≥ P0.99). Data were stored and analyzed using the According to the K-L grading scale, 35 patients were K-L
MILLIPLEX® Analyst 5.1. 2, 65 patients were K-L 3, and 48 patients were K-L 4.There
The study was approved by the ethics committee of the were no significant differences in age, weight, height, and
Nanjing Drum Tower Hospital, The Affiliated Hospital of BMI among the KOA patients. The serum levels of leptin
Nanjing University Medical School. (K-L 2, 34,260.5 ± 4527.3 pg/ml) and OPN (K-L 2, 5684.1
± 727.1 pg/ml) in the K-L 2 group were higher than those of
Statistical analysis K-L 3 and K-L 4, and there were significant differences be-
tween K-L 2 group and K-L 4 group in leptin; in addition,
The data were statistically analyzed by using SPSS 16.0 soft- there were significant differences in OPN between K-L 2
ware (IBM, Armonk, NY, USA). Besides, the results were group and K-L 3 group, and the serum SOST (K-L 4,
expressed as mean ± standard deviation (SD) or mean (range). 2711.6 ± 678.6 pg/ml) level in K-L 4 group was remarkably
The differences in serum levels of leptin, OPN, and SOST higher than that in K-L 2 group (Table 1).
between the groups were compared by using chi-square test As shown in Table 2, further analysis of data from women
for categorical variables, Student’s t test for normally and men, the female KOA patients’ age, weight, and BMI
 Clin Rheumatol

Table 1 Comparison of serum biomarkers in the KOA patients and control subjects

Control (n = 101) KOA patients

Total (n = 148) K-L 2 (n = 35) K-L 3(n = 65) K-L 4(n = 48) p value

Sex(M/F %) 45/56 36/112 7/30 12/51 7/41

Age(y) 57.7(41–79) 68.0(43–87)# 67.4(59-80) 67.8(44–84) 68.5(43–87) ns
Weight(kg) 65.5(42.7–92.1) 68.6(43–105)# 72.1(43-87) 67.9(46–105) 69.7(55–85) ns
Height(m) 1.63(1.46–1.8) 1.59(1.5–1.74) 1.59(1.5–1.70) 1.60(1.5–1.71) 1.59(1.5–1.74) ns
BMI (kg/m2) 24.2 ± 3.0 25.5 ± 7.2 26.4 ± 4.2 26.3 ± 3.7 24.3 ± 3.9 ns
Leptin (pg/ml) 6936.4 ± 702.2 26,581.7 ± 2011.5## 34,260.5 ± 4527.3* 26,908.6 ± 2727.9 23,884.5 ± 6883.2* 0.038
OPN (pg/ml) 2182.5 ± 217.8 4908.3 ± 769.4## 5684.1 ± 727.1* 4494.6 ± 592.5* 5136.0 ± 673.3 0.043
SOST (pg/ml) 1288.9 ± 267.7 2481.9 ± 543.5## 2205.6 ± 849.7* 2390.5 ± 579.2 2711.6 ± 678.6* 0.04
#
p < 0.05 (vs control)
##
p < 0.01 (vs control)
*p < 0.05
**p < 0.01
Normally distributed data are expressed as mean ± SD. Data that are abnormally distributed are shown as mean (range)

were significantly higher than those in non-KOA females There were same situation in KOA group but no significant
(p < 0.05), while this was not observed in men. Additionally, difference.
the serum levels of leptin, OPN, and SOST were markedly
higher in male and female KOA patients than those in the Correlation between the severity of KOA imaging and
control group. There were significant differences in the three SOST, OPN, and leptin
bone turnover factors between the KOA group and the control
group. Serum levels of Leptin in control group, female’s As demonstrated in Table 3, there was a significant correlation
higher than male (F,8529.4 ± 591.7pg/ml vs M,6235.9 ± between bone turnover factor and K-L grading as follows:
729.4 p < 0.05); OPN in control group male’s higher than OPN and SOST (r = 0.211, p = 0.013) and SOST and K-L
male (M, 2539.9 ± 566.1 pg/ml vs F, 1632.0 ± 226.7 pg/ml). (r = 0.193, p = 0.035). There was a correlation between the
following bone turnover factors in control group: height and
leptin (r = 0.249, p = 017), BMI and leptin (r = −0.229, p =
Table 2 Differences in serum biomarkers between the KOA patients
and control subjects, gender separated 027), and OPN and SOST (r = 0.405, p = 0.0001).

Male Control (n = 45) KOA patients (n = 36) p value

Incidence analysis and ROC curve analysis
Age (year) 57.5(45–79) 72.3(55–82) p < 0.05
Weight (kg) 72.2(49–92) 73.7(51–95) ns As presented in Table 4, univariate regression analysis of
Height (m) 1.70(1.57–1.80) 1.68(1.6–1.74) ns data demonstrated that leptin and OPN were not remark-
BMI (kg/m2) 24.3 ± 2.7 25.7 ± 4.4 ns ably associated with an increased risk of OA (OR = 1,
Leptin (pg/ml) 6235.9 ± 729.4 23,003.1 ± 2400.1 p < 0.001 p < 0.05; OR = 1, p < 0.05) in all the study subjects. The
OPN (pg/ml) 2539.9 ± 566.1 5538.1 ± 723.0 p < 0.05 serum levels of SOST were positively associated with
SOST (pg/ml) 1354.2 ± 380.3 2540.5 ± 403.6 p < 0.05 incidence of OA. The risk of incident KOA increased
Female Control (n = 56) KOA patients (n = 112) p value with elevation of serum levels of SOST (OR = 1.001,
Age (year) 57.7(41–75) 67.0(43–87) p < 0.05 p < 0.001). Similarly, analysis of data revealed that
Weight(kg) 60.1(43–83) 67.6(43–105) p < 0.05 SOST were important risk factors for KOA, while the
Height(m) 1.57(1.46–1.68) 1.58(1.5–1.71) ns odds ratios (ORs) were not remarkable. After adjustment
BMI (kg/m2) 24.2 ± 3.2 26.3 ± 6.4 p < 0.05 for age, sex, height, and BMI, multivariate regression
Leptin (pg/ml) 8529.4 ± 591.7 26,802.1 ± 4077.2 p < 0.001 analysis of data uncovered that serum levels of SOST
OPN (pg/ml) 1632.0 ± 226.7 4545.8 ± 556.3 p < 0.001
were significantly correlated with incidence of KOA
SOST (pg/ml) 960.1 ± 104.6 2280.4 ± 160.6 p < 0.05
(OR = 1.001, p < 0.001). Similarly, analysis of data
disclosed that SOST in females were the most important
Age, weight, height, mean (range); BMI, leptin, OPN, SOST, mean ± SD risk factors for OA (OR = 1.001, p < 0.05).
Clin Rheumatol

Table 3 Correlation of serum

leptin, osteopontin, and sclerostin KOA patients Leptin OPN SOST K-L
levels and clinical parameters in
all study subjects Age r = 0.118p = 0.192 r = 0.011p = 0.907 r = 0.094p = 0.299 r = 0.43 p = 0.655
Weight r = 0.064p = 0.527 r = 0.036p = 0.725 r = −0.04p = 0.732 r = −0.018 p = 0.867
Height r = −0.11p = 0.337 r = 0.136p = 0.228 r = 0.053p = 0.638 r = −0.007p = 0.953
BMI r = 0.142p = 0.096 r = 0.015p = 0.858 r = 0.071p = 0.408 r = −0.105 p = 0.244
Leptin r = −0.007p = 0.933 r = −0.046p = 0.597 r = −0.062 p = 0.503
OPN r = 0.211p = 0.013* r = 0.045. p = 0.636
SOST r = 0193. p = 0.035*
Control Leptin OPN SOST K-L
Age r = 0.086p = 0.40 r = 0.075p = 0.463 r = 0.115p = 0.261 ..
Weight r = −0.175p = 0.97 r = −0.019p = 0.859 r = 0.043p = 0.684 ..
Height r = 0.249p = 0.17 r = −0.081p = 0.444 r = −0.087p = 0.407 ..
BMI r = −0.229p = 0.27 r = −0.146p = 0.163 r = 0.095p = 0.360 ..
Leptin r = 0.189p = 0.063 r = −0.068p = 0.504 ..
OPN r = 0.405p = 0.0001** ..

The table shows Pearson and spearman correlations coefficients (R) in four groups

As illustrated in Fig. 1, the ROC curves revealed a Discussion

relationship between serum levels of leptin, OPN, and
SOST and incidence of KOA. Furthermore, the analysis Our study demonstrates elevated levels of leptin, OPN, SOST,
showed the highest values of area under ROC (AUROC) age, and weight in KOA patients group. Unlike our study, in
curve for serum levels of leptin, OPN, and SOST (leptin, the Carrie et al. study, after adjusting for BMI, serum leptin
0.829, 95% confidence interval (CI) 0.776–0.883; OPN, level increased with age and increased by 0.38 ng/ml per year
0.698, 95% CI 0.631–0.765; SOST, 0.748, 95% CI on average [17], suggesting that our population differ with
0.686–0.810). With respect to the cutoff values calculated respect to body size. Serum leptin levels in human blood were
from the ROC curves, the serum levels of leptin, OPN, related to sex and age [18]. Therefore, the reasons for the
and SOST showed sensitivity of 87.2%, 54.1% and 60.9% inconsistencies with other studies may be the sample sex ratio
and specificity of 68.4%, 77.6%, and 77.6%, respectively and race. Serum OPN [19] and SOST [20] levels had a signif-
(cutoff values, 5529, 2036.5, and 1733.5 pg/ml, icant positive correlation with age, and OPN [18] had a neg-
respectively). ative correlation with body weight, and height. OPN and
SOST levels may be influenced by BMI and height in our
results (Table 3).

Leptin

As we all know, obesity is one of the main risk factors for

weight-bearing and non-heavy joint osteoarthritis [21]. At
present, in obese OA patients, leptin has no clear consensus.
Researchers found that there is a genetic correlation of leptin
and OA [22]. Therefore, further research is needed for a large
number of patients to clarify and provide new information
about the correlation between leptin and OA. In the serum of
patients with OA, leptin has been always found to be in-
creased [23, 24]. Zhang et al. [25] demonstrated that leptin is
closely correlated with obesity and could be an important risk
factor for OA in Asian, Caucasians, and mixed populations.
Dumond et al. [26] found that leptin can intensively enhance
the metabolism of chondrocyte and is associated with the le-
Fig. 1 The receiver operating characteristic curve generated from the
sion of the knee. Scotece et al. achieved a similar conclusion
leptin, OPN, and SOST serum levels measured in KOA and without [27]. Carrie et al. carried out a 10-year follow-up, which indi-
KOA cated that both baseline and value after follow-up in OA group
 Clin Rheumatol

Table 4 The variables associated

with incidence of KOA by Univariate analysis Multivariate analysis
univariable and multivariable
analysis M and F OR (95%CI) p value OR (95%CI) p value
Leptin 1 (1,1) 0.0001 1 (1,1) 0.0001
OPN 1 (1,1) 0.0001 1 (1,1) 0.0001
SOST 1.001(1.001, 1.001) 0.0001 1.001 (1.000,1.001) 0.0001
Male
Leptin 1 (1,1) 0.003 1 (11001) 0.058
OPN 1 (11001) 0.026 1 (11001) 0.078
SOST 1.001 (1.001,1.002) 0.001 1.002 (0.999,1004) 0.156
Female
Leptin 1 (1,1) 0.0001 1 (1,1) 0.0001
OPN 1 (1,1) 0.001 1 (11001) 0.003
SOST 1.001 (1.001,1.001) 0.0001 1.001 (1.000,1.001) 0.001

were notably beyond those of control group [18]. In the pres- positively related to the severity of OA [30]. Sittisak et al.
ent study, we found that serum levels of leptin increased in and Gao SG et al. achieved the same result [31, 32].
OA patients compared with non-KOA subjects, indicating that Although circulating and/or synovial fluid levels of OPN have
alterations in leptin expression may be related to OA; thus, the been studied in patients with knee osteoarthritis, there have
high serum level of leptin may be a risk factor for OA. been no detailed studies of plasma and synovial fluid levels of
Furthermore, our findings highlighted that leptin level was OPN at different clinical stages of primary knee osteoarthritis.
higher in female OA patients than male OA patients, and there The present study highlighted that OPN concentration in
were similar differences in the healthy subjects, suggesting KOA group was significantly higher than that in normal con-
significant differences in gender among the groups. trol group (KOA, 4908.3 ± 769.4 pg/ml, vs control, 2182.5 ±
Consistent with Carrie et al.’s results [18], in the present study, 217.8 pg/ml). In a previous study grouped by gender, serum
we also observed that the expression level of leptin was cor- OPN level was the highest in male KOA patients, while that
related with the pathogenesis of OA in different genders, and was the lowest in healthy female. Our results are in agreement
the high level of leptin was particularly important as a poten- with the findings of Honsawek et al. [31] and Mohammed
tial risk factor in female OA patients. Estradiol may enhance et al. [31] who performed similar studies on patients with
leptin production in women, while testosterone may inhibit KOA and noted that those male KOA patients had the highest
leptin production in men. Leptin might be helpful in the iden- plasma OPN level compared with the lowest in non-KOA
tification of subgroups of OA patients with high risk. Several female (KOA male, 5538.1 ± 723.0 pg/ml, vs control female,
previously conducted studies reported that leptin is associated 1632.0 ± 226.7 pg/ml). The findings of the present research
with OA and incidence of OA. The ROC curve analysis indicate that the serum OPN level in KOA patients with K-L 2
showed that leptin had a remarkable diagnostic value in the (5684.1 ± 727.1 pg/ml) was greater than that with K-L grade 3
incidence of KOA. To sum up, leptin levels are associated (4494.6 ± 592.5 pg/ml), and serum OPN level was the highest
with OA progression, suggesting a possible role as a good in K-L 2 patients in the KOA group, followed by K-L 4;
biomarker for monitoring disease progression and diagnosis. increase of serum level of OPN is not associated with progres-
sive grades of advanced KOA, while that does not necessarily
OPN mean that OPN is not associated with pro-inflammatory or
anti-inflammatory factors. According to ROC curve analysis,
OPN could upregulate expression levels of interleukin 6 (IL- serum OPN level showed sensitivity of 54.1% and specificity
6) and interleukin 8 (IL-8) in human chondrocytes OA, and of 77.6%, and the cutoff value of serum OPN level between
the increased expression could increase concentration of healthy individuals and patients with KOA was ≥ 2036.5 pg/
OPN, which might be one of the potential mechanisms of ml. In the present study, it was disclosed that there were pos-
OPN in the OA [28]. In articular cartilage, the correlation itive and negative correlations between OPN and SOST and
between Wnt5a and OPN might be important to the develop- PTH, respectively, in the KOA patients (r = 0.211, p = 0.013;
ment and pathogenesis of KOA [29]. The level of OPN in OA r = −0.188, p = 0.038). There was a positive correlation be-
plasma and synovial fluid is remarkably higher and is tween OPN and SOST in control group (r = 0.405, p =
Clin Rheumatol

0.0001). This may be related to potential diseases associated In the current research, all the patients with autoimmune
with risk factors (e.g., aging and obesity) for OA. diseases, chronic or acute inflammation, or cancers were ex-
cluded to rule out the possibility of a change in the patients’
levels of bone turnover biomarkers, while there were poten-
SOST tially chronic diseases that could affect the results. This was a
retrospective study; therefore, the results would be unavoid-
To date, existing studies have comprehensively investigated ably affected by sampling error. Secondly, the sample size
serum levels of SOST in OA patients. Serum levels of SOST was limited; therefore, further studies need to be conducted
depend on genetic aspects, as well as age, sex, adiposity, kid- to achieve a more persuasive conclusion. In addition, we only
ney function, and presence of diabetes mellitus. Testing of focused on serum instead of simultaneous analysis of plasma
patients with KOA confirmed that they had significantly and synovial fluid. Furthermore, in the absence of cytology
higher SOST levels in all of the measured serum levels com- and functional genomics experiments, this research can mere-
pared with the control group. Mo¨ dder et al. al’ data demon- ly reflect the superficial phenomenon, and the research direc-
strated that the positive relationship between age and tion can be followed by a series of in vitro and in vivo
sclerostin levels in both genders [33]. Our data did not show experiments.
the same trend. This may be related to the difference in sample In conclusion, in this cross-sectional study consisting of
age selection. In the KOA group, serum SOST concentration 249 serum samples, we demonstrated higher serum levels of
increased with K-L grading(r = 0193. p = 0.035). In contrast SOST, OPN, and leptin compared with non-KOA controls.
to the results of the present study, Mabey et al. demonstrated Additionally, we noted a strong association between grades
that serum SOST level was negatively correlated with severity of SOST and K-L, which appeared to be a novel finding. Our
of KOA and reported that it might play a protective role results may contribute to the identification of novel targets for
against OA [7]. In the progression of osteoarthritis, from ini- medical interventions and also assist scholars in the future
tial cartilage destruction to eventual subchondral researches concentrating on bone markers.
osteosclerosis and massive osteophyte proliferation, during
this process, intra-articular hyperosteogeny is active, leading Funding information This work was supported by the National Science
Foundation of China (81672239).
to an increase in SOST reactivity that inhibits bone formation
through intervention in the Wnt signaling pathway [34].
Although OA serum levels of sclerostin changing appear to Compliance with ethical standards
be associated with the formation of bony spurs, we do not
Disclosures None.
currently know whether it actually causes syndesmophyte for-
mation or whether it is a response to it. On the other hand, in Ethical approval All procedures performed in studies involving human
mice models, there is a dose-dependent decrease in sclerostin participants were in accordance with the ethical standards of the regional
expression with mechanical loading, whereas unloading in- research committee (Regional Ethical Review Board for CFDA-GCP/
creases SOST expression [35, 36].Therefore, we can infer that ICH-GCP) and with the 1964 Helsinki declaration and its later amend-
ments or comparable ethical standards.
as osteoarthritis becomes more severe, joint load gradually
decreases, and SOST expression increases, and serum level
increases. This inference needs further experimental proof.
Other important finding of our study is that the serum level
References
of SOST in men was significantly higher than that in women
in the KOA group and the non-KOA control group (KOA, 1. Felson DT, Dsc YZ, Hannan MT, Naimark A, Levy D (1995) The
2540.5 ± 403.6 pg/ml vs 2280.4 ± 160.6 pg/ml; control, incidence and natural history of knee osteoarthritis in the elderly.
1354.2 ± 380.3 pg/ml vs 960.1 ± 104.6 pg/ml).This gender The Framingham osteoarthritis study. Arthritis Rheum 38(10):
difference is consistent with the previous studies [37]. Many 1500–1505
2. Leung GJ, Rainsford KD, Kean WF (2014) Osteoarthritis of the
scholars evaluated whether genetic loss or pharmacologic in- hand I: aetiology and pathogenesis, risk factors, investigation and
hibition of SOST has an impact on knee joint articular carti- diagnosis. J Pharm Pharmacol 66(3):339–346
lage. Differences in disease progression, population, methods 3. Garnero P (2001) Cross sectional evaluation of biochemical
of analysis, and incomplete control of confounding variables markers of bone, cartilage, and synovial tissue metabolism in pa-
(e.g., physical activity and eating) were in disagreement with tients with knee osteoarthritis: relations with disease activity and
joint damage. Ann Rheum Dis 60(6):619–626
the findings of the present study. According to the ROC curve 4. Atkins GJ, Rowe PS, Lim HP, Welldon KJ, Ormsby R, Wijenayaka
analysis, among the studied markers, the serum SOST level AR et al (2011) Sclerostin is a locally acting regulator of late-oste-
could predict OA and its severity. Biochemical markers of oblast/preosteocyte differentiation and regulates mineralization
SOST may facilitate early diagnosis, as well as monitoring through a mepe-asarm-dependent mechanism. J Bone Miner Res
26(7)
the responsiveness of treatments.
 Clin Rheumatol

5. Blom AB, Peter LVL, Peter MVDK, Wim BVDB (2010) To seek possible molecular target for osteoarthritis therapeutic intervention.
shelter from the wnt in osteoarthritis? Wnt-signaling as a target for Ann Rheum Dis 66(12):1616–1621
osteoarthritis therapy. Curr Drug Targets 11(5):620–629 23. Masuzaki H, Ogawa Y, Isse N, Satoh N, Okazaki T, Shigemoto M,
6. Bouaziz W, Funck-Brentano T, Lin H, Marty C, Ea HK, Hay E, Mori K, Tamura N, Hosoda K, Yoshimasa Y, Jingami H, Kawada
Cohen-Solal M (2015) Loss of sclerostin promotes osteoarthritis in T, Nakao K (1995) Human obese gene expression: adipocyte-
mice via β-catenin-dependent and -independent wnt pathways. specific expression and regional differences in the adipose tissue.
Arthritis Res Ther 17(1):24 Diabetes 44(7):855–858
7. Mabey T, Honsawek S, Tanavalee A, Wilairatana V, 24. Pelleymounter MA, Cullen MJ, Baker MB, Hecht R, Winters D,
Yuktanandana P, Saetan N, Zhan D (2014) Plasma and synovial Boone T (1995) Effects of the obese gene product on body weight
fluid sclerostin are inversely associated with radiographic severity regulation in Ob/Ob mice. Science 269(5223):540–543
of knee osteoarthritis. Clin Biochem 47(7–8):547–551. https://doi. 25. Zhang P, Zhong ZH, Yu HT, Liu B (2015) Significance of in-
org/10.1016/j.clinbiochem.2014.03.011 creased leptin expression in osteoarthritis patients. PLoS One:10
8. Denhardt DT, Noda M (1998) Osteopontin expression and func- 26. Dumond H, Presle N, Terlain B, Mainard D, Loeuille D, Netter P
tion: role in bone remodeling. 72(Supplement 30):92–102 et al (2014) Evidence for a key role of leptin in osteoarthritis.
9. Attur MG, Dave MN, Stuchin S, Kowalski AJ, Steiner G, Arthritis Rheum 48(11):3118–3129
Abramson SB, Denhardt DT, Amin AR (2001) Osteopontin: an 27. Scotece M, Mobasheri A et al (2015) Leptin in osteoarthritis: focus
intrinsic inhibitor of inflammation in cartilage. Arthritis Rheum on articular cartilage and chondrocytes. Life Sci 140:
44(3):578–584 S0024320515003161
10. Pullig O, Weseloh G, Gauer S, Swoboda B (2000) Osteopontin is 28. Yang Y, Gao SG, Zhang FJ, Luo W, Lei GH (2014) Effects of
expressed by adult human osteoarthritic chondrocytes: protein and osteopontin on the expression of il-6 and il-8 inflammatory factors
mRNA analysis of normal and osteoarthritic cartilage. Matrix Biol J in human knee osteoarthritis chondrocytes. Cell Biochem Biophys
Int Soc Matrix Biol 19(3):245–255 29. Yusheng L, Wenfeng X, Minghua S, Zhenhan D, Chao Z, Hui L
11. Green ED, Maffei M, Braden VV, Proenca R, Desilva U, Zhang Y et al (2016) The expression of osteopontin and wnt5a in articular
et al (1995) The human obese (Ob) gene: RNA expression pattern cartilage of patients with knee osteoarthritis and its correlation with
and mapping on the physical, cytogenetic, and genetic maps of disease severity. Biomed Res Int 2016:1–7
chromosome 7. Genome Res 5(1):5–12 30. Mohammed FI, Abd El-Azeem MI, Kamaleldin AM (2012) Plasma
12. Cowley MA, Smart JL, Rubinstein M, Cerdan MG, Diano S, and synovial fluid osteopontin levels in patients with knee osteoar-
Horvath TL et al (2001) Leptin activates anorexigenic pomc neu- thritis: relation to radiological grade. Egypt Rheumatol 34(3):131–
rons through a neural network in the arcuate nucleus. Nature 136
411(6836):480–484
31. Honsawek S, Tanavalee A, Sakdinakiattikoon M, Chayanupatkul
13. Liang J, Feng J, Wu WKK, Xiao J, Wu Z, Han D, Zhu Y, Qiu G
M, Yuktanandana P (2009) Correlation of plasma and synovial
(2011) Leptin-mediated cytoskeletal remodeling in chondrocytes
fluid osteopontin with disease severity in knee osteoarthritis. Clin
occurs via the rhoa/rock pathway. J Orthop Res 29(3):369–374
Biochem 42(9):808–812
14. Scotece M, Conde J, López V, Lago F, Pino J, Gómez-Reino JJ,
32. Gao SG, Li KH, Zeng KB, Tu M, Xu M, Lei GH (2010) Elevated
Gualillo O (2014) Adiponectin and leptin: new targets in inflam-
osteopontin level of synovial fluid and articular cartilage is associ-
mation. Basic Clin Pharmacol Toxicol 114(1):97–102
ated with disease severity in knee osteoarthritis patients. Osteoarthr
15. Altman RD, Asch E, Bloch DA, Bole G, Hochberg M (1986)
Cartil 18(1):82–87
Development of criteria for the classification and reporting of oste-
33. Mödder UI, Hoey KA, Amin S, Mccready LK, Achenbach SJ,
oarthritis. Classification of osteoarthritis of the knee. Diagnostic
Riggs BL et al (2011) Relation of age, gender, and bone mass to
and therapeutic criteria committee of the American rheumatism
circulating sclerostin levels in women and men. J Bone Miner Res
association. Arthritis Rheum 29(8):1039–1049
26(2):373–379
16. Kellgren JH, Lawrence JS (1957) Radiological assessment of
34. Semenov M, Tamai K, He X (2005) SOST is a ligand for lrp5/lrp6
osteoarthrosis. Ann Rheum Dis 16(4):494–502
and a wnt signaling inhibitor. J Biol Chem 280(29):26770–26775
17. Karvonen-Gutierrez CA, Harlow SD, Mancuso P, Jacobson J,
Mendes de Leon CF, Nan B (2013) Association of leptin levels 35. Moustafa A, Sugiyama T, Prasad J, Zaman G, Gross TS, Lanyon
with radiographic knee osteoarthritis among a cohort of midlife LE, Price JS (2012) Mechanical loading-related changes in osteo-
women. Arthritis Care Res 65(6):936–944 cyte sclerostin expression in mice are more closely associated with
18. Ahima RS, Prabakaran D, Mantzoros C, Qu D, Lowell B, Maratos- the subsequent osteogenic response than the peak strains engen-
Flier E, Flier JS (1996) Role of leptin in the neuroendocrine re- dered. Osteoporos Int 23(4):1225–1234
sponse to fasting. Nature (London) 382(6588):250–252 36. Robling AGA, Niziolek PJP, Baldridge LAL, Condon KWK, Allen
19. Chang IC, Chiang TI, Yeh KT, Lee H, Cheng YW (2010) Increased MRM, Alam II et al (2008) Mechanical stimulation of bone in vivo
serum osteopontin is a risk factor for osteoporosis in menopausal reduces osteocyte expression of SOST/sclerostin. J Biol Chem
women. Osteoporos Int 21(8):1401–1409 283(9):5866–5875
20. Ardawi M-SM, Al-Kadi HA, Rouzi AA et al (2011) Determinants 37. Wanby P, Nobin R, Von SP, Brudin L, Carlsson M (2016) Serum
of serum sclerostin in healthy pre- and postmenopausal women. J levels of the bone turnover markers dickkopf-1, sclerostin, osteo-
Bone Miner Res 26(12) protegerin, osteopontin, osteocalcin and 25-hydroxyvitamin d in
21. Cicuttini FM, Baker JR, Spector TD (1996) The association of swedish geriatric patients aged 75?Years or older with a fresh hip
obesity with osteoarthritis of the hand and knee in women: a twin fracture and in healthy controls. J Endocrinol Investig 39:855–863
study. J Rheumatol 23:1221–1226
22. Iliopoulos D, Malizos KN, Tsezou A (2007) Epigenetic regulation Publisher’s note Springer Nature remains neutral with regard to jurisdic-
of leptin affects mmp-13 expression in osteoarthritic chondrocytes: tional claims in published maps and institutional affiliations.