Scandinavian Journal of Clinical & Laboratory Investigation, 2011; 71: 631–636

ORIGINAL ARTICLE

Serum osteocalcin levels in patients with nonalcoholic fatty

liver disease: Association with ballooning degeneration

YUSUF YILMAZ1,2, RAMAZAN KURT1,2, FATIH EREN2 & NESE IMERYUZ1,2

1Department
of Gastroenterology, Marmara University, School of Medicine, Istanbul, and 2Institute of Gastroenterology,
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Waterloo on 11/20/14

Marmara University, Maltepe, Istanbul, Turkey

Abstract
Our aim was to examine the relation of serum osteocalcin (OCN) levels with the clinical, biochemical, and histological
characteristics of patients with biopsy-proven nonalcoholic fatty liver disease (NAFLD). We carried out a case-control study
including 99 patients with biopsy-proven NAFLD and 75 age- and sex-matched controls. Concentrations of OCN were
measured in aprotinin-treated serum samples using a solid-phase enzyme amplified sensitivity immunoassay. Serum OCN
levels were significantly lower in patients with NAFLD than in healthy controls. In patients with NAFLD, serum OCN
levels were inversely associated with ALT (r  0.36, p  0.001), AST (r  0.39, p  0.001), HOMA-IR (r  0.30,
p  0.01) and the degree of hepatocyte ballooning (r  0.20, p  0.05). Serum OCN was the only independent predictor
of the degree of hepatocyte ballooning in NAFLD patients (β  0.24; t  2.146, p  0.05). Compared with controls,
For personal use only.

NAFLD patients have a decrease in serum OCN concentrations, which is significantly associated with serum transaminases
and the extent of hepatocyte ballooning.

Key Words: Osteocalcin, nonalcoholic fatty liver disease, hepatocyte ballooning

Introduction
recently shown to act as a bone-derived hormone in
Nonalcoholic fatty liver disease (NAFLD), a hepatic the regulation of energy metabolism [8,9]. Com-
manifestation of the metabolic syndrome [1], is the pared to wild-type animals, OCN -/- knockout mice
most common liver disease in developed countries are characterized by an abnormal amount of fat
and an important cause of abnormal liver function mass, an increase in serum triglycerides, impaired
tests in hepatology practice [2]. The term NAFLD is insulin secretion, insulin resistance, and glucose
used to describe a wide spectrum of fatty liver changes intolerance [10]. Interestingly, the expression of
ranging from simple steatosis to nonalcoholic steato- insulin target genes in the liver appears to be uni-
hepatitis (NASH) [3]. Accumulating research sug- formly decreased in OCN-null mice [10]. Clinical
gests that NAFLD is independently associated with human studies have reported a significant inverse
insulin resistance [4]. In addition, insulin resistance association between OCN and insulin resistance,
may predict the severity of liver damage or fatty infil- blood glucose, adiposity, and triglycerides [11–15].
tration in patients with NAFLD [5]. These observa- Furthermore, recent work has demonstrated a
tions indicate that either insulin resistance plays a negative relationship between serum OCN and the
role in the pathogenesis and progression of liver presence of the metabolic syndrome in different eth-
damage, or the two phenomena have a common nic groups [12,15]. In the setting of chronic liver
pathogenic mechanism [6]. diseases, a pilot study reported decreased serum
Osteocalcin (OCN), the most abundant non- OCN levels in patients with primary biliary cirrhosis
collagenous protein found in bone, is produced by and in those with chronic alcoholic liver disease.
osteoblasts [7]. Besides its role as a marker of bone In a recent study of 28 obese patients, Fernández-
formation and bone turnover [7], OCN has been Real et al. [17] have shown that circulating OCN

Correspondence: Yusuf Yilmaz, MD, Institute of Gastroenterology, Marmara University, P.K. 53, Basibuyuk, Maltepe, Istanbul 34840, Turkey. Tel: 90
5334403995. Fax: 90 2166886681. E-mail: [email protected]

(Received 14 May 2011; accepted 1 July 2011)

ISSN 0036-5513 print/ISSN 1502-7686 online © 2011 Informa Healthcare
DOI: 10.3109/00365513.2011.604427
 632 Y. Yilmaz et al.

concentrations are negatively associated with The Ethics Committee of the Marmara Univer-
blood markers of liver injury and liver disease, includ- sity School of Medicine approved this study and
ing alanine transaminase (ALT) and aspartate all participants provided written informed consent
transaminase (AST). In addition, the changes in prior to participation.
ALT levels following weight loss in obese individuals
were linearly associated with changes in OCN Clinical and biochemical characterization
concentrations [17].
Because currently the information on the All subjects underwent physical examination, anthro-
relationships between OCN and NAFLD is lacking, pometric measurements and biochemical screening.
in this study we sought to examine the nature Body mass index (BMI) was calculated from
and the strength of the associations between serum measurements of height and weight. Diabetes mel-
concentrations of OCN and the severity of liver litus was diagnosed according to ADA criteria [18].
histology among patients with biopsy-proven The metabolic syndrome was diagnosed using the
NAFLD. Clarification of these associations may be ATP III criteria [19]. The estimate of insulin resis-
of clinical importance in planning preventative and tance was calculated using the HOMA-IR index,
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Waterloo on 11/20/14

therapeutic strategies. with the following formula: insulin resistance 

fasting plasma insulin (in microunits per milliliter)
FPG (in millimoles per liter)/22.5. Blood pressure
Subjects and methods was measured using a mercury sphygmomanometer
Study participants in a quiet room after  10-min rest. Korotkoff 1
and 5 were taken for systolic blood pressure and dia-
The study consists of 99 patients with biopsy- stolic blood pressure, respectively. Routine blood
proven NAFLD (50 males and 49 females, mean chemistry analyses were performed at the central
age, 48  8 years) and 75 healthy subjects (37 males laboratory of clinical chemistry of our hospital.
and 38 females, mean age, 48  7 years). The two Serum high-sensitivity C-reactive protein (hs-CRP)
groups were comparable in age and sex. The selec- levels were measured in duplicate, random order,
For personal use only.

tion criteria for biopsy were the following: (i) fatty and in a blinded fashion using a commercially avail-
liver on ultrasound with persistent elevations of liver able method (Dade Behring, Marburg, Germany).
function tests for 6 months; (ii) fatty liver on ultra- The intra-assay and the inter-assay coefficients of
sound with hepatomegaly and/or splenomegaly even variation for hs-CRP were 4.6% and 6.1%, respec-
in the absence of elevations of liver function tests. tively, and the lower detection limit was 0.12 mg/L.
Patients with NAFLD were consecutively seen at our
hospital-based specialized outpatient clinics over the
Liver histology
past 12 months. During the enrollment period, a
total of 110 patients met the selection criteria for Ultrasonography-guided liver biopsies were per-
biopsy. However, 10 subjects refused biopsy and formed under conscious sedation using a 16-gauge
one was excluded because of a concomitant diagno- Hepafix needle. All biopsy specimens were placed
sis of colon cancer. Therefore, 99 NAFLD patients in formalin solution for fixation and embedded in
were included in the final analysis. All showed paraffin blocks. Serial sections (sectioned at 4 mm
ultrasonographic evidence of steatosis grade 1 or intervals) were stained with hematoxylin-eosin,
higher. Patients with viral hepatitis, hemochromato- Masson’s trichrome. An experienced pathologist
sis, Wilson’s disease, autoimmune hepatitis, primary blinded to clinical data scored the liver biopsies
biliary cirrhosis, sclerosing cholangitis, biliary according to the NIDDK NASH Clinical Research
obstruction, alpha-1 antitrypsin deficiency, ischemic Network scoring system [20]. Steatosis was scored
cardiac or cerebrovascular disease, impaired renal from 0–3 with a four-grade scoring system from
function, or malignancies were carefully excluded S0–S3: S0: no steatosis or less than 5%, S1: 5–33%,
from the present study. Subjects using estrogens, S2: 33–66%, S3:  66%. Lobular inflammation was
amiodarone, steroids, tamoxifen, and lipid lowering graded as follows: stage 0, no foci; stage 1:  2 foci
agents were not eligible for this study. Patients per 200 field; stage 2: 2–4 foci per 200 field; stage
with daily alcohol intake exceeding 20 g/day or previ- 3:  4 foci per 200 field. Ballooning degeneration
ous intestinal surgery were also excluded. A total of of liver cells was evaluated as: grade 0, absent; grade
75 healthy age- and gender-matched volunteers 1, few cells; grade 2, many cells. The histological
served as controls. All subjects included in the control NASH score was defined as the unweighted sum of
group were judged to be in good health, with normal the scores for steatosis (0–3), lobular inflammation
results on liver function tests and confirmed as hav- (0–3), and ballooning (0–2); thus ranging from 0–8.
ing normal liver by ultrasound. Subjects with a con- Cases with scores of 0–2 were considered as having
sumption of alcohol  20 g/day or who were taking simple steatosis; on the other hand, cases with scores
any medication were not included in the control group. of 5 or greater were diagnosed as definitive NASH.
All patients and controls were of Turkish descent. Cases with activity scores of 3 and 4 were considered
 Serum osteocalcin in NAFLD 633

as borderline NASH [20]. Fibrosis was staged as their natural units for presentation in the text and
follows: stage 0: no fibrosis; stage 1: perisinusoidal tables. The Student’s t-test was used to evaluate dif-
or periportal fibrosis with three different patterns: ferences between the two study groups in normally
1A: mild, zone 3, perisinusoidal; 1B: moderate, zone distributed continuous variables. When normality
3, perisinusoidal fibrosis, and 1C portal/periportal was not confirmed, the Mann–Whitney U test was
fibrosis; stage 2: perisinusoidal and portal/periportal used. Correlations among the study variables were
fibrosis; stage 3: bridging fibrosis; stage 4: cirrhosis. tested by the Spearman’s correlation coefficient.
Multivariable stepwise linear regression analyses
were performed to identify independent predictors
Measurement of serum OCN levels of the severity of histological features of NAFLD (i.e.
All blood samples were collected from an antecubital steatosis, lobular inflammation, hepatocye balloon-
vein between 8:00 and 9.00 a.m. after fasting ing, liver fibrosis); the covariates included in these
overnight. Samples were centrifuged at 2500 g for models were OCN and all variables are listed in
10 min and aprotonin (100 μL) was added to all Table I. All statistical analyses were performed using
SPSS version 11.0 for Windows (SPSS, Inc.,
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Waterloo on 11/20/14

samples immediately after centrifugation. Serum ali-

quots were then stored at 80°C until immediately Chicago, IL, USA). Statistical significance was
before analysis. Serum OCN levels were determined defined at p  0.05 using two-sided tests. The
using a solid-phase enzyme-amplified sensitivity Bonferroni’s correction was not applied due to
immunoassay performed on microtiter plates the exploratory nature of the study and the prefer-
(GenWay hOST-EASIA, GenWay Biotech, Inc., San ence to reduce the likelihood of Type II error.
Diego, CA, USA) according to the manufacturer’s
protocol. This assay uses monoclonal antibodies Results
directed against distinct epitopes of human osteocal-
cin. Calibrators and samples react with the capture Table I displays the general characteristics of subjects
monoclonal antibody coated on microtiter well and with and without NAFLD. The two study groups
did not differ in terms of age, gender, systolic and
For personal use only.

with a monoclonal antibody labelled with horserad-

ish peroxidase. After an incubation period allowing diastolic blood pressure, and HDL cholesterol.
the formation of a sandwich, the microtiterplate Between-group comparison analysis identified a
was washed to remove unbound enzyme-labelled number of variables to be different in patients with
antibody. Bound enzyme-labelled antibody was histology-proven NAFLD compared with controls
measured through a chromogenic reaction. The without NAFLD, including body mass index,
amount of substrate turnover was determined colori- HOMA-IR, AST, ALT, total cholesterol, LDL cho-
metrically by measuring the absorbance, which lesterol, and tryglicerides. The prevalence of diabetes
is proportional to the osteocalcin concentration. All and the metabolic syndrome was higher in patients
measurements were performed in duplicate, in with NAFLD than in controls.
random order, and the results were averaged. All Serum OCN levels were significantly lower in
determinations were performed in a blinded fashion. patients with NAFLD (median 2.22 ng/mL; inter-
The detection limit, defined as the apparent concen- quartile range: 2.15–2.41 ng/mL) than in healthy
tration two standard deviations above the average controls (median 2.30 ng/mL; interquartile range:
optical density at zero binding, was 0.4 ng/mL. 2.22–2.42 ng/mL, Mann–Whitney U test, p  0.001,
The intra- and inter-assay coefficients of variance Figure 1). The group of NAFLD patients included
were 4.5% and 3.3%, respectively. four subjects with an exceptionally high serum
OCN level (3.50 ng/mL). No obvious explanation
of these high levels was evident. However, even after
Data analysis
exclusion of these outliers, serum OCN levels
The study power was calculated using the StatMate in NAFLD patients (median 2.21 ng/mL; inter-
software, version 2.0 (GraphPad, San Diego, CA, quartile range: 2.15–2.37 ng/mL) remained signifi-
USA). Our sample had a 90% power to detect, cantly lower compared with controls. We then
between patients with NAFLD and controls, a dif- evaluated the association between serum OCN
ference of 0.21 ng/mL in serum OCN levels with a levels and the clinical, biochemical, and histological
significance level (alpha) of 0.05 (two-sided). The phenotypes in our sample. In patients with NAFLD,
Kolmogorov–Smirnov test was performed in all con- serum OCN levels were inversely associated with
tinuous variables to define the presence of normality. ALT (r  0.36, p  0.001), AST (r  0.39,
Gaussian variables are expressed as mean  standard p  0.001), HOMA-IR (r  0.30, p  0.01) and the
deviation (SD), skewed data are reported as medians degree of hepatocyte ballooning (r  0.20, p  0.05).
and interquartile ranges, and categorical variables are There was no association between serum OCN levels
expressed as counts. Skewed variables were logarith- and the NASH score (r  0.08, p  0.41), and this
mically transformed to improve normality prior to molecule did not discriminate between simple steato-
statistical analysis and then back-transformed to sis (n  50; median 2.23 ng/mL; interquartile range:
 634 Y. Yilmaz et al.

Table I. General characteristics of the study participants.

NAFLD patients (n  99) Control group (n  75) p value

Sex (M/F) 50/49 37/38 NS

Age (years) 48  8 48  7 NS
Body mass index (kg/m2) 30.6  4.9 27.4  4.3 0.01
Diabetes mellitus (yes/no) 29/70 0/75 0.001
Metabolic syndrome (yes/no) 61/38 0/75  0.001
HOMA-IR 3.7 (2.35.1) 1.6 (0.62.5) <0.001
Systolic blood pressure (mmHg) 138  22 126  20 NS
Diastolic blood pressure (mmHg) 85  13 82  11 NS
AST (U/L) 44  18 24  10 0.001
ALT (U/L) 68  32 21  11 0.001
Total cholesterol (mmol/L) 5.6  1.3 4.9  1.2 0.01
HDL cholesterol (mmol/L) 1.2  0.3 1.1  0.3 NS
LDL cholesterol (mmol/L) 3.8  1.3 3.3  0.6 0.001
Triglycerides (mmol/L) 2.1  0.9 1.6  0.8 0.001
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Waterloo on 11/20/14

hs-CRP (mg/L) 3.2 (2.64.9) − −

Histological steatosis 2 (13) − −
Lobular inflammation 2 (13) − −
Ballooning 2 (12) − −
NASH score 5 (47) − −
Fibrosis 1 (02) − −

HOMA-IR, homeostasis model of insulin resistance; AST, aspartate aminotransferase; ALT, alanine aminotransferase; HDL, high-density
lipoprotein; LDL, low-density lipoprotein; hs-CRP, high sensitivity C-reactive protein; NASH, nonalcoholic steatohepatitis. Data are
presented as means and SD, counts, or medians and interquartile ranges, as appropriate.

2.14–2.42 ng/mL) and definite NASH (n  49; Discussion

For personal use only.

median 2.21 ng/mL; interquartile range: 2.16–2.30

In the present report, we describe the association
ng/mL). We did not identify other characteristics
between serum OCN, clinical characteristics and
associated with serum OCN levels in our patients
liver histology in patients with biopsy-proven NAFLD.
with NAFLD.
This study has shown, for the first time, that (i)
After stepwise linear regression analysis adjusting
patients with biopsy-proven NAFLD have significantly
for age, sex, body mass index, diabetes mellitus, met-
lower serum OCN concentrations than matched
abolic syndrome, HOMA-IR, blood pressure values,
controls; (ii) OCN concentrations are inversely asso-
liver enzymes, lipid variables, and hs-CRP, serum
ciated with HOMA-IR, AST, and ALT in patients
OCN levels retained their independent significance
with NAFLD; and (iii) OCN concentrations are
as a predictor of the degree of hepatocyte ballooning
weakly and inversely associated with hepatocyte bal-
in patients with NAFLD (β  0.24; t  2.146,
looning among NAFLD patients after adjustment for
p  0.05).
a broad spectrum of potential confounders, includ-
ing age, sex, body mass index, diabetes mellitus,
metabolic syndrome, HOMA-IR, blood pressure
values, liver enzymes, lipid variables, and hs-CRP.
As observed in the non-NAFLD setting, OCN
levels correlated strongly with HOMA-IR [15], AST
and ALT [17] even in our NAFLD patients. Another
interesting finding is the weak but significant nega-
tive correlation of OCN with the degree of hepato-
cyte ballooning, which to our knowledge has not
been previously reported. Of note, after adjustment
for potential confounders including HOMA-IR,
OCN remained negatively associated with hepato-
cyte ballooning. The negative relationship between
HOMA-IR and OCN observed in our study was pre-
viously reported [15] and further suggests a bone-
metabolism cross-talk in the regulation of insulin
resistance [21]. The strong inverse relation we found
Figure 1. Scatter diagram for serum OCN levels in NAFLD between OCN and serum transaminases was previ-
patients and healthy controls. Horizontal lines across the scatter ously described in obese individuals [17]. We found
diagram represent mean values. in the present study that this relation is also present
 Serum osteocalcin in NAFLD 635

in patients with biopsy-proven NAFLD, suggesting in serum OCN concentrations (vs. matched controls)
that, independent of increased BMI, other not yet which are weakly but significantly associated with
identified factors linked to direct hepatocyte injury the extent of hepatocyte ballooning, independent of
may explain this increase. Since in this study OCN other risk factors (including insulin resistance and
levels were inversely associated with hepatocyte bal- the metabolic syndrome). To confirm our observa-
looning, independent of insulin resistance and other tions, larger validation analyses and longitudinal
features of the metabolic syndrome, it could also be prospective studies are necessary.
hypothesized that this molecule might play a role in
the development and progression of NAFLD. How-
ever, there was only a weak correlation between the Acknowledgements
two measures, which suggests that caution needs This study was supported by grants from the Marmara
to be exercised when drawing pathophysiological University Research Fund (SAG-C-TUP-090909-
links. Hepatocyte ballooning indicates hepatocyte 0274) and the Turkish Association for the Study
degeneration associated with enlargement, swelling, of Liver Diseases.
rounding, and characteristic reticulated cytoplasm.
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Waterloo on 11/20/14

Ballooning is considered a structural manifestation

of microtubular disruption and is likely a represen- Declaration of interest: The authors report no
tation of cells undergoing lytic necrosis [22,23]. conflicts of interest. The authors alone are respon-
As serum AST and ALT levels are conventionally sible for the content and writing of the paper.
believed to be surrogate biomarkers of hepatocyte
injury and/or death, the associations observed in
this study seem to suggest that serum OCN levels References
may possibly reflect liver injury or lytic necrosis in [1] Erickson SK. Nonalcoholic fatty liver disease. J Lipid Res
particular. Clearly, we must be cautious in making 2009;50:S412–6.
any causal inference, given both the weak correla- [2] Sass DA, Chang P, Chopra KB. Nonalcoholic fatty liver
disease: a clinical review. Dig Dis Sci 2005;50:171–80.
tion between OCN and hepatocyte ballooning
For personal use only.

[3] Almeda-Valdés P, Cuevas-Ramos D, Aguilar-Salinas CA.

and the cross-sectional design of this study. However, Metabolic syndrome and non-alcoholic fatty liver disease.
our current results clearly indicate that hepatocyte Ann Hepatol 2009;8:S18–24.
ballooning – rather than other pathological findings [4] Byrne CD, Olufadi R, Bruce KD, Cagampang FR,
– represents the main histological correlate of reduced Ahmed MH. Metabolic disturbances in non-alcoholic fatty
liver disease. Clin Sci (Lond) 2009;116:539–64.
OCN levels in patients with NAFLD.
[5] Grønbaek H, Thomsen KL, Rungby J, Schmitz O,
Several limitations of this study need to be Vilstrup H. Role of nonalcoholic fatty liver disease in the
addressed. First, this study was not designed to development of insulin resistance and diabetes. Expert Rev
address the mechanisms behind the observed asso- Gastroenterol Hepatol 2008;2:705–11.
ciations. Second, the relatively small sample size lim- [6] Yilmaz Y, Ulukaya E, Gul OO, Arabul M, Gul CB, Atug O,
Oral AY, Aker S, Dolar E. Decreased plasma levels of
its the generalizability of our conclusions. Third, the
soluble receptor for advanced glycation endproducts
study population consisted of Turkish men and (sRAGE) in patients with nonalcoholic fatty liver disease.
women; as a result, it is uncertain whether these find- Clin Biochem 2009;42:802–7.
ings are generalized in other ethnic groups. Finally, [7] Lee AJ, Hodges S, Eastell R. Measurement of osteocalcin.
the assay used in this study measures both carboxy- Ann Clin Biochem 2000;37:432–46.
[8] Kim YS, Paik IY, Rhie YJ, Suh SH. Integrative physiology:
lated and uncarboxylated forms of OCN. Human
defined novel metabolic roles of osteocalcin. J Korean
studies have suggested that elevated levels of both Med Sci 2010;25:985–91.
carboxylated and uncarboxylated forms of OCN are [9] Lieben L, Callewaert F, Bouillon R. Bone and metabolism:
associated with improved glucose tolerance [24], a complex crosstalk. Horm Res 2009;71(Suppl. 1):134–8.
while the carboxylated form is mainly associated with [10] Lee NK, Sowa H, Hinoi E, Ferron M, Ahn JD, Confavreux
C, Dacquin R, Mee PJ, McKee MD, Jung DY, Zhang Z,
insulin resistance [25]. Our study measured total
Kim JK, Mauvais-Jarvis F, Ducy P, Karsenty G. Endocrine
OCN levels and did not have measurements of uncar- regulation of energy metabolism by the skeleton. Cell
boxylated OCN. Future studies are needed to mea- 2007;130:456–69.
sure uncarboxylated and carboxylated forms of OCN [11] Fernández-Real JM, Izquierdo M, Ortega F, Gorostiaga E,
and to determine their association with histology Gómez-Ambrosi J, Moreno-Navarrete JM, Frühbeck G,
Martínez C, Idoate F, Salvador J, Forga L, Ricart W,
findings in patients with NAFLD. Despite these lim-
Ibañez J. The relationship of serum osteocalcin concentra-
itations, to our knowledge, this is the first study to tion to insulin secretion, sensitivity, and disposal with
demonstrate the relationship between OCN and liver hypocaloric diet and resistance training. J Clin Endocrinol
injury in patients with NAFLD from the aspects of Metab 2009;94:237–45.
both clinical and pathological findings. In addition, [12] Pittas AG, Harris SS, Eliades M, Stark P, Dawson-Hughes
B. Association between serum osteocalcin and markers
all patients with NAFLD underwent liver biopsy, the
of metabolic phenotype. J Clin Endocrinol Metab 2009;94:
best standard for diagnosing this condition [26,27]. 827–32.
In conclusion, our results show that patients with [13] Kindblom JM, Ohlsson C, Ljunggren O, Karlsson MK,
biopsy-proven NAFLD have significant reductions Tivesten A, Smith U, Mellström D. Plasma osteocalcin
 636 Y. Yilmaz et al.

is inversely related to fat mass and plasma glucose in MS, Unalp-Arida A, Yeh M, McCullough AJ, Sanyal AJ;
elderly Swedish men. J Bone Miner Res 2009;24:785–91. Nonalcoholic Steatohepatitis Clinical Research Network.
[14] Kanazawa I, Yamaguchi T, Yamamoto M, Yamauchi M, Design and validation of a histological scoring system
Kurioka S, Yano S, Sugimoto T. Serum osteocalcin level is for nonalcoholic fatty liver disease. Hepatology 2005;41:
associated with glucose metabolism and atherosclerosis 1313–21.
parameters in type 2 diabetes mellitus. J Clin Endocrinol [21] Kassi E, Papavassiliou AG. A possible role of osteocalcin in
Metab 2009;94:45–9. the regulation of insulin secretion: human in vivo evidence?
[15] Saleem U, Mosley Jr TH, Kullo IJ. Serum osteocalcin is J Endocrinol 2008;199:151–3.
associated with measures of insulin resistance, adipokine [22] Fujii H, Ikura Y, Arimoto J, Sugioka K, Iezzoni JC, Park SH,
levels, and the presence of metabolic syndrome. Arterioscler Naruko T, Itabe H, Kawada N, Caldwell SH, Ueda M.
Thromb Vasc Biol 2010;30:1474–8. Expression of perilipin and adipophilin in nonalcoholic fatty
[16] Szalay F, Lakatos P, Németh J, Abonyi M, Büki B, Tarján G, liver disease; relevance to oxidative injury and hepatocyte
Holló I. [Decreased serum osteocalcin level in non-alcoholic ballooning. J Atheroscler Thromb 2009;16:893–901.
and alcoholic chronic liver diseases]. Orv Hetil 1991;132: [23] Caldwell S, Ikura Y, Dias D, Isomoto K, Yabu A, Moskaluk
1301–5. C, Pramoonjago P, Simmons W, Scruggs H, Rosenbaum N,
[17] Fernández-Real JM, Ortega F, Gómez-Ambrosi J, Wilkinson T, Toms P, Argo CK, Al-Osaimi AM, Redick JA.
Salvador J, Frühbeck G, Ricart W. Circulating osteocalcin Hepatocellular ballooning in NASH. J Hepatol 2010;53:
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Waterloo on 11/20/14

concentrations are associated with parameters of liver fat 719–23.

infiltration and increase in parallel to decreased liver enzymes [24] Hwang YC, Jeong IK, Ahn KJ, Chung HY. The uncarboxy-
after weight loss. Osteoporos Int 2010;21:2101–7. lated form of osteocalcin is associated with improved glucose
[18] ACE/ADA Task Force on Inpatient Diabetes. American tolerance and enhanced β-cell function in middle-aged
College of Endocrinology and American Diabetes Associa- male subjects. Diabetes Metab Res Rev 2009;25:768–72.
tion Consensus statement on inpatient diabetes and glyce- [25] Shea MK, Gundberg CM, Meigs JB, Dallal GE,
mic control. Diabetes Care 2006;29:1955–62. Saltzman E, Yoshida M, Jacques PF, Booth SL. Gamma-
[19] Grundy SM, Brewer HB Jr, Cleeman JI, Smith SC Jr, carboxylation of osteocalcin and insulin resistance in older
Lenfant C; American Heart Association; National Heart, men and women. Am J Clin Nutr 2009;90:1230–5.
Lung, and Blood Institute. Definition of metabolic syndrome: [26] Emanuele E. Is biopsy always necessary? Toward a
Report of the National Heart, Lung, and Blood Institute/ clinico-laboratory approach for diagnosing nonalcoholic
American Heart Association conference on scientific issues steatohepatitis in obesity. Hepatology 2008;48:2086–7.
related to definition. Circulation 2004;109:433–8. [27] Yilmaz Y, Dolar E. Biomarkers for noninvasive biochemical
[20] Kleiner DE, Brunt EM, Van Natta M, Behling C, diagnosis of nonalcoholic steatohepatitis: tools or decorations?
For personal use only.

Contos MJ, Cummings OW, Ferrell LD, Liu YC, Torbenson World J Gastroenterol 2009;15:4346–7.