Sage Journals: Discover world-class research

Abstract

Background

The association of plasma cardiovascular risk markers and metabolic syndrome (MetS) with non-alcoholic fatty liver disease (NAFLD) has not been well defined.

Methods

Japanese men (n = 809) had standard anthropometric measurements done, and had their liver fat quantitated by ultrasound. Three groups were identified: (1) normal controls without significant disease, (2) preliminary-metabolic syndrome (pre-MetS) cases and (3) MetS cases. Plasma adiponectin, high sensitivity-C reactive protein (hs-CRP), HOMA-IR, lipids, lipoproteins and liver enzymes were evaluated among the three groups.

Results

The prevalence of fatty liver was 13% in controls, 39% in pre-MetS and 62% in MetS. Plasma adiponectin and high density lipoprotein cholesterol (HDL-C) were significantly decreased, and HOMA-IR, hs-CRP, TG, remnant lipoproteins (RLPs) and small dense-LDL-C (sd LDL-C) were significantly increased in subjects with fatty liver compared to those without fatty liver. Multivariate analyses of serum parameters associated with fatty liver revealed that adiponectin and hs-CRP were more strongly associated with the presence of fatty liver than waist circumference. However, HOMA-IR, HDL-C, TG, RLP-C, RLP-TG and sd LDL-C were more strongly associated with waist circumference than with fatty liver. Factor analysis revealed that adiponectin and HDL-C were linked to liver enzymes, lipoproteins and HOMA-IR associated with fatty liver, but not with waist circumference.

Conclusions

Adiponectin was found to be a more specific diagnostic marker for the presence of fatty liver regardless of MetS status, and was inversely correlated with liver enzyme concentrations. However, RLPs were found to be more specifically associated with the presence of MetS.

Keywords

Metabolic syndrome non-alcoholic fatty liver adiponectin high sensitivity-C reactive protein remnant lipoprotein-cholesterol remnant lipoprotein-triglyceride

Introduction

Non-alcoholic fatty liver disease (NAFLD) is estimated to affect about 30% of the general population and over 70% of patients with type 2 diabetes.¹ Given the expected continuing increase in the prevalence of both obesity and diabetes, NAFLD will soon become an epidemic, if it is not one already. The consequences of NAFLD are numerous and range from chronic liver disease to a worsening of insulin resistance and type 2 diabetes, to being a contributor to both cardiovascular disease (CVD) and chronic kidney disease (CKD).² NAFLD is, therefore, a complex problem with implications far beyond the liver. It is well known that metabolic syndrome (MetS) leads to CVD, mainly due to a number of metabolic and inflammatory alterations.³ In addition, MetS is closely associated with NAFLD, which comprises the hepatic manifestation of this syndrome.⁴

Plasma adiponectin concentration has been proposed as a diagnostic marker of MetS from many clinical studies.^5–8 A reduced plasma concentration of adiponectin has been said to be associated with increased visceral fat. However, we have often diagnosed cases of MetS without low plasma adiponectin concentrations. Furthermore, there have been cases without MetS, but with significantly decreased plasma adiponectin concentrations. Therefore, we have investigated cases with decreased plasma adiponectin concentrations, which may be more closely associated with other conditions than MetS. Our goal in this study was to evaluate cases with fatty liver in pathophysiology associated with the expression of increased visceral fat, most of which is diagnosed as MetS or preliminary-MetS (pre-MetS). We also evaluated the other plasma markers associated with fatty liver such as C-reactive protein (CRP), HOMA-IR, high density lipoprotein cholesterol (HDL-C), remnant lipoproteins (RLP), small dense LDL-cholesterol (sd LDL-C) and liver enzymes in this study. Those diagnostic markers are also known as the markers of CVDs. The diagnosis of fatty liver was made using ultrasonography, a non-invasive, inexpensive method of moderate specificity and sensitivity.⁹ Although fatty liver has been considered to be a part of MetS, a recent Framingham Heart Study report indicated that fatty liver is a prevalent condition, characterized by dysglycemia and dyslipidemia, which in fact is independent from the visceral adipose tissue shown to be a major source of adiponectin secretion.¹⁰

Adipokine alterations, such as reduced adiponectin and elevated tumour necrosis factor alpha (TNF-α) and interleukin-6 (IL-6) are considered the key factors in the development of the insulin resistance underlying MetS, contributing to the pro-inflammatory state.¹¹ Reduced adiponectin and elevated TNF-α and IL-6 may also induce the secretion of CRP in liver and an excessive intrahepatocyte free fatty acid pool promoting the deposition of triglycerides (TGs) in the hepatic parenchyma and favouring fatty liver,¹² although it is still controversial whether this is dependent on or independent of insulin resistance. Also, it has been not clear yet, whether the circulating adipokines and CRP concentrations are different in MetS with and without hepatic steatosis, and through their presence, generating a more pronounced chronic inflammatory state.¹³ In addition, adhesion molecules, crucial for leukocyte interaction with the endothelium and migration into the artery wall, are reportedly elevated in patients with MetS.¹⁴ Elevated plasma RLP was reported to induce plasma vascular cell adhesion molecule (sVCAM-1) and intracellular adhesion molecule (sICAM-1), which is implicated in the endothelial inflammatory and atherogenic process in MetS patients.^15,16

We evaluated adiponectin, CRP, insulin resistance (HOMA-IR) and lipoproteins along with currently used diagnostic markers for liver injuries, comparing those with and without fatty liver in the healthy controls, pre-MetS and MetS cases. It has already been reported that there is overproduction and secretion of very low density lipoproteins (VLDL) in insulin resistance associated with NAFLD, irrespective of the presence of liver fat deposits.¹⁷ Adiponectin regulates lipid metabolism and may influence VLDL production and/or catabolism which determines VLDL-apoB concentration and plasma TG concentrations.¹⁸ It is not clear if the adiponectin in MetS has a different role in VLDL when a fatty liver is present. Plasma RLP cholesterol (RLP-C), triglyceride (RLP-TG) and RLP-TG/RLP-C ratio (which reflects the particle size) were also compared in MetS with and without fatty liver in terms of parameters of lipoprotein metabolism.

The aim of this study was to clarify whether the plasma markers such as adiponectin, hs-CRP, HOMA-IR and lipoproteins could be the diagnostic marker of NAFLD associated with a risk of CVD or a marker of MetS. We have also discussed the risk of CVD and its relationship with NAFLD associated with hypoadiponectinaemia in this manuscript.

Materials and methods

Subjects

The study cases were all men because the prevalence of MetS in Japan using Japanese criteria is very low in women.¹⁹ In our analysis, a total of 809 men underwent screening who were not receiving medication for hypertension, diabetes or lipid abnormalities. In this population, we defined the MetS participants as those having abdominal obesity (85 cm or greater in waist circumference) and two or more of the following components: (1) a systolic blood pressureof 130 mmHg or higher or a diastolic blood pressure of 85 mmHg or higher, (2) a serum TG concentration of 1.7 mmol/L (150 mg/dL) or higher and HDL-C <1.0 mmol/L (40 mg/dL) and (3) a fasting plasma glucose of 6.1 mmol/L (110 mg/dL) or higher. One hundred and twenty three subjects met these criteria for MetS. In addition, we screened the subjects for meeting the criteria for pre-MetS (pre-MetS), defined as those having a waist circumference of 85 cm or greater, but fewer than two components of MetS described above. A total of 305 subjects met the criteria for pre-MetS. In addition, 382 normal control subjects were identified who had a waist circumference of <85 cm, and who had no components for MetS as described above. These three groups were selected to provide a continuum ranging from normal to pre-MetS and MetS. All these subjects were identified while receiving a health check-up at Hidaka Hospital, Takasaki, Japan from April 2008 to October 2009, and were randomly selected for the present study. Subjects having the following were excluded: (1) alcohol intake >20 g/day (certified by the frequency to drink alcohol), (2) recent history of acute illness, (3) taking medications known to modify lipid concentrations, blood pressure or blood glucose, (4) taking medications known to induce a fatty liver, (5) being seropositive for hepatitis B or C and/or (6) having any other health problem known to be associated with liver steatosis. In all the cases men were 35–65 years old. The cases of fatty liver in this study were identified as non-alcoholic fatty liver (NAFLD).

All the patients were referred to the haepatology unit at Hidaka Hospital for ultrasound investigation of hepatic steatosis, performed as a part of routine clinical practice. According to the ultrasound results, subjects in the groups with MetS, pre-MetS and controls were further divided into those with and without a fatty liver. The fatty liver positive group included the cases who presented with ‘bright liver’ echo patterns according to standard criteria (i.e. evidence of a diffuse increase in echogenicity of the liver compared with that of the kidneys). The non-fatty liver group with MetS, pre-MetS and normal controls comprised of the cases who did not exhibit any echogenic evidence of fatty liver. The weight and height of each participant were measured and body mass index (BMI) was calculated to evaluate for the presence of obesity. Waist circumference measurement was made midway between the lateral lower rib margin and the superior anterior iliac crest in a standing position in all cases by the same investigator. Written informed consent was obtained from all of the participants. The study had the approval of the Ethical Committees of Hidaka Hospital and of Gunma University School of Medicine.

Biochemical analysis

After 12-hour overnight fast, blood samples were taken, and sera were isolated and stored at 4℃. The following analytes were measured within 48 h of drawing the blood: liver enzymes, haemoglobin A1c (HbA1c), total cholesterol (TC), TGs, direct low density lipoprotein cholesterol (LDL-C) and high density lipoproteins (HDL-C) as previously reported.²⁰ Aliquots of serum were also stored at −80℃ for subsequent determination of adiponectin, high sensitivity C-reactive protein (hs-CRP), RLP-C and RLP-TG. Specifically, the reagent used for measuring LDL-C and HDL-C was a direct method obtained from Denka Seiken (Tokyo, Japan) and hs-CRP was a latex-based immunoturbidometric assay obtained from Denka Seiken. In addition, the adiponectin assay was also a latex-based immunoturbidometric method obtained from Otsuka (Tokyo, Japan) and the immunoseparation methods of RLP-C and RLP-TG (JIMRO II) were obtained from Otsuka. Automated sd LDL-C assay kit was provided from Denka-Seiken. Liver enzymes including (aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma-glutamyltranspeptidase (GGT) as markers of hepatic damages and glucose, insulin and HbA1c were measured using commercial enzymatic kits on a Hitachi 917 auto-analyzer. In order to estimate the insulin resistance, HOMA-IR index (a homeostatic model for the assessment for insulin resistance) was calculated as fasting insulin (μmol/mL) × fasting glucose (mmol/L)/22.5 at BML (Tokyo).

Statistical analysis

The data are presented as median values with 25th and 75th percentile values, rather than as mean values with standard deviations because almost all variables, especially markers of inflammation, were not normally distributed. Accordingly, differences between groups were tested by the Mann-Whitney U test and the following multivariate analyses were done after logarithmic transformation of test results for AST, ALT, GGT, TG, RLP-C, RLP-TG, sd LDL-C, adiponectin, hs-CRP, HbA1c and HOMA-IR. Multiple regression analysis was performed for each variable with age, presence of fatty liver (binary: 0 or 1) and waist circumference (cm) set as explanatory variables in order to examine the relative contribution of ‘fatty liver’ to the waist circumference adjusted for age. On the other hand, the relative contribution of the variables in diagnosing the status of fatty liver was evaluated by use of multiple logistic regression analysis. In addition, factor analysis was performed to explore the commonality among the variables based on the principal component method with the varimax rotation of the factorial axes. Comparison of factor scores derived from the analysis between patients with or without fatty liver was made by use of receiver operating characteristic (ROC) analysis. The area under the ROC curve was used as an index of accuracy in differentiating the two conditions. All these statistical analyses were performed by use of StatFlex Ver 6 (Artech, Osaka, Japan). P values <0.05 were considered statistically significant.

Results

The clinical and biochemical characteristics in controls, pre-MetS and MetS cases

Table 1 shows the clinical and biochemical characteristics of the controls, pre-MetS and MetS cases in this study group. Each group was further subdivided into those with or without the presence of fatty liver as determined by ultrasound (Table 2). Significant differences in BMI and waist circumference were found among the three groups, as well as between the cases with fatty liver and non-fatty liver in each group (Table 2). The most significant difference among the three groups was the prevalence of fatty liver, with 13% of the controls, 39% of the pre-MetS group and 62% of the MetS group having this abnormality. These data indicated a 4.8-fold higher prevalence of fatty liver in MetS than in controls. Parameters of lipid metabolism, liver function, glucose homeostasis and inflammation were all significantly different in MetS when compared to controls. Subjects in the pre-MetS group generally had values for all these parameters that were intermediate between these two groups (Table 1).

Table 1.

Clinical and biochemical characteristics of metabolic syndrome and control subjects in men.

	Normal (A) n = 382	Pre-metabolic syndrome (B) n = 305	Metabolic syndrome (C) n = 122	P
	Median (25th percentile–75th percentile)	Median (25th percentile –75th percentile)	Median (25th percentile –75th percentile)	A vs. B	A vs. C	B vs. C
Age (years)	48 (43–55)	50 (44–56)	51 (46–56)	0.740	0.228	0.537
Body weight (kg)	61.6 (58.1–65.5)	74 (69.1–79.7)	76 (71–84)	<0.001	<0.001	0.003
BMI (kg/cm²)	21.6 (20.6–22.8)	25 (23.9–26.8)	26 (24.3–28.2)	<0.001	<0.001	<0.001
Waist circumference (cm)	80 (76–82)	90 (87–94)	91 (88–99)	<0.001	<0.001	<0.001
Fatty liver(%)	13	39	62	<0.001	<0.001	<0.001
Blood pressure
Systolic (mmHg)	116 (105–125)	118 (111–126)	130 (130–139)	0.034	<0.001	<0.001
Diastolic (mmHg)	74 (66–82)	78 (72–83)	87 (82–94)	0.001	<0.001	<0.001
TC (mmol/L)	5.1 (4.6–5.7)	5.1 (4.7–5.6)	5.4 (4.7–6.0)	1.000	0.004	0.006
TG (mmol/L)	1.1 (0.8–1.5)	1.3 (0.9–1.6)	1.9 (1.4–2.4)	0.177	<0.001	<0.001
HDL-C (mmol/L)	1.5 (1.2–1.8)	1.3 (1.2–1.5)	1.3 (1.1–1.5)	<0.001	<0.001	0.518
LDL-C (mmol/L)	3 (2.6–3.6)	3.2 (2.7–3.7)	3.3 (2.7–4.5)	0.001	<0.001	0.227
sd-LDL (mmol/L)	0.73 (0.56–1.04)	0.84 (0.72–1.04)	46.5 (32.6–56.6)	0.206	<0.001	<0.001
RLP-C (mmol/L)	0.11 (0.08–0.15)	0.12 (0.09–0.17)	0.21 (0.15–0.31)	0.815	<0.001	<0.001
RLP-TG (mmol/L)	0.09 (0.11–0.34)	0.12 (0.07–0.22)	0.29 (0.18–0.51)	0.859	<0.001	<0.001
RLP-TG/RLP-C	0.76 (0.51–1.18)	1.03 (0.66–1.44)	1.42 (1.03–1.86)	0.003	<0.001	<0.001
RLP-TG/TG	0.08 (0.06–0.12)	0.1 (0.07–0.15)	0.15 (0.12–0.23)	0.052	<0.001	<0.001
HbA1c (%)	5.1 (4.9–5.3)	5.2 (5.0–5.3)	5.4 (5.2–6.2)	0.374	<0.001	<0.001
HOMA-IR	0.9 (0.7–1.3)	1.4 (1.0–2.0)	1.7 (1.2–2.9)	0.029	0.028	0.939
AST (IU/L)	21 (18–25)	22 (19–26)	24 (20–30)	0.317	<0.001	0.001
ALT (IU/L)	20 (16–27)	25 (19–35)	28 (22–44)	<0.001	<0.001	0.006
ALP (IU/L)	202 (171–238)	202 (171–234)	195 (173–235)	0.955	0.986	1.000
GGT (IU/L)	28 (19–46)	36 (25–52)	48 (22–75)	0.178	<0.001	<0.001
ChE (IU/L)	340 (303–386)	363 (325–401)	369 (336–426)	0.003	<0.001	0.072
Albumin (g/dL)	4.4 (4.2–4.6)	4.4 (4.2–4.5)	4.4 (4.2–4.5)	0.026	0.304	0.976
hs-CRP (mg/L)	0.30 (0.20–0.70)	0.40 (0.20–0.73)	0.60 (0.40–1.20)	0.200	0.593	0.762
Adiponectin (µmol/mL)	7.7 (5.8–9.9)	6.2 (4.9–8.3)	5.8 (4.5–7.7)	0.020	<0.001	0.166

Table 2.

Plasma concentrations of adiponectin, hs-CRP, HDL-C, HOMA-IR and remnant lipoproteins in subjects with and without metabolic syndrome according to the status of fatty liver.

Normal			Pre-metabolic syndrome			Metabolic syndrome
	Non-fatty liver (n = 333)	Fatty liver (n = 49)	P	Non-fatty liver (n = 185)	Fatty liver (n = 120)	P	Non-fatty liver (n = 46)	Fatty liver (n = 76)	P
	Median (25th percentile –75th percentile)	Median (25th percentile –75th percentile)	P	Median (25th percentile –75th percentile)	Median (25%tile–75%tile)	P	Median (25%tile–75%tile)	Median (25th percentile –75th percentile)	P
Age (year)	48 (43–55)	49 (44–55)	0.725	51 (44–56)	49 (43–55)	0.081	51 (47–56)	51 (45–55)	0.495
Body weight (kg)	61.2 (57.8–65.2)	63.7 (60.3–66.6)	0.004	72.9 (68.4–77.9)	77.1 (70.7–82.9)	<0.001	71.5 (67.3–76.4)	78.6 (73.0–87.7)	<0.001
BMI	21.4 (20.4–22.7)	22.5 (21.6–23.8)	<0.001	24.8 (23.7–26.0)	26 (24.5–28.5)	<0.001	24.3 (22.6–27.0)	26.6 (25.4–29.7)	<0.001
Waist circumference (cm)	79 (75–82)	82 (80–83)	<0.001	89 (87–93)	92 (89–97)	<0.001	88 (87–92)	94 (89–101)	0.001
Blood pressure (mmHg)
Systolic	116 (106–124)	116 (105–127)	0.784	119 (111–127)	117 (111–124)	0.199	132 (127–141)	129 (123–137)	0.038
Diastolic	74 (66–81)	77 (67–84)	0.364	78 (72–84)	78 (72–82)	0.556	90 (85–96)	86 (81–92)	0.042
TC (mmol/L)	5.1 (4.6–5.7)	5.2 (4.7–5.7)	0.720	5.1 (4.7–5.5)	5.2 (4.7–5.7)	0.130	5.7 (4.9–6.0)	5.4 (4.6–6.0)	0.380
TG (mmol/L)	1.1 (0.8–1.4)	1.5 (1.1–2.2)	<0.001	1.2 (0.9–1.5)	1.5 (1.1–1.9)	<0.001	2.0 (1.6–2.6)	1.8 (1.4–2.4)	0.333
HDL-C (mmol/L)	1.5 (1.3–1.8)	1.3 (1.2–1.5)	<0.001	1.4 (1.2–1.6)	1.2 (1.1–1.4)	<0.001	1.4 (1.2–1.7)	1.2 (1.0 – 1.3)	<0.001
LDL-C (mmol/L)	3.0 (2.6–3.5)	3.2 (2.8–3.7)	0.109	3.1 (2.7–3.6)	3.4 (2.9–3.9)	0.002	3.2 (2.6–3.8)	3.4 (2.7–4.1)	0.163
sd LDL-C (mmol/L)	0.69 (0.54–1.00)	0.97 (0.66–1.10)	0.227	0.83 (0.65–0.98)	0.90 (0.77–1.16)	0.077	1.15 (0.98–1.47)	1.21 (0.79–1.46)	0.606
RLP-C (mmol/L)	0.11 (0.08–0.14)	0.14 (0.09–0.23)	0.060	0.11 (0.08–0.16)	0.13 (0.10–0.22)	0.054	0.20 (0.15–0.34)	0.21 (0.16–0.29)	0.950
RLP-TG (mmol/L)	0.08 (0.04–0.13)	0.15 (0.08–0.33)	0.013	0.11 (0.06–0.19)	0.17 (0.05–0.25)	0.001	0.29 (0.19–0.50)	0.29 (0.18–0.51)	0.728
RLP-TG/RLP-C	0.74 (0.50–1.12)	1.15 (0.90–1.30)	0.013	0.91 (0.58–1.29)	1.27 (0.95–1.63)	0.001	1.43 (1.06–1.90)	1.42 (1.01–1.79)	0.538
RLP-TG/TG	0.08 (0.06–0.11)	0.12 (0.09–0.16)	0.010	0.09 (0.06–0.14)	0.11 (0.09–0.16)	0.005	0.15 (0.11 – 0.23)	0.15 (0.12–0.24)	0.884
HbA1c (%)	5.1 (4.9–5.3)	5.2 (5.0–5.4)	0.014	5.1 (5.0–5.3)	5.2 (5.1–5.5)	0.001	5.2 (5.1–5.4)	5.7 (5.4–6.8)	<0.001
HOMA-IR	0.9 (0.7–1.3)	1.3 (1.1–1.7)	0.154	1.1 (0.8–1.8)	1.9 (1.5–2.6)	<0.001	1.2 (0.9–1.6)	2.5 (1.7–3.9)	<0.001
AST (IU/L)	21 (18–24)	23 (20–26)	0.011	21 (18–24)	25 (21–32)	<0.001	23 (20 – 27)	25 (21–34)	0.091
ALT (IU/L)	19 (16–25)	28 (22–35)	<0.001	21 (17–28)	34 (25–32)	<0.001	25 (18–28)	35 (25–52)	<0.001
ALP (IU/L)	199 (170–236)	220 (182–261)	0.007	197 (171–229)	208 (172–248)	0.104	193 (174–234)	200 (165–234)	0.876
GGT (IU/L)	26 (19–43)	34 (28–52)	0.001	33 (23–49)	42 (29–54)	0.005	51 (32–79)	45 (33–73)	0.866
ChE (IU/L)	337 (300–377)	370 (324–417)	0.002	348 (317–385)	381 (349–422)	<0.001	355 (330–398)	376 (343–434)	0.037
Albumin (g/dL)	4.4 (4.2–4.5)	4.5 (4.3–4.6)	0.101	4.4 (4.3–4.5)	4.4 (4.3–4.5)	0.004	4.3 (4.2–4.4)	4.4 (4.2–4.5)	0.025
hs-CRP (mg/L)	0.30 (0.15–0.65)	0.60 (0.50–0.95)	0.010	0.30 (0.20–0.60)	0.60 (0.30–1.23)	0.003	0.50 (5.4–8.5)	0.70 (0.50–1.40)	0.028
Adiponectin (mg/L)	8.0 (6.3–10.2)	5.2 (3.9–7.6)	<0.001	7.0 (5.1–8.7)	5.4 (4.4–7.1)	0.001	7.1 (5.4–8.5)	5.1 (4.3–6.5)	<0.001

The comparison of diagnostic parameters among the cases with or without fatty liver in controls, pre-MetS and MetS cases

In Table 2, the data are presented as median values with 25th and 75th percentile values, rather than as mean values with standard deviations because almost all variables, especially markers of inflammation, were not normally distributed. All of the parameters among the three groups are further divided into those with or without a fatty liver. These results show the stronger associations with lipoproteins (elevated TG, RLP-TG, RLP-C/RLP-TG and low HDL-C), markers of glucose homeostasis (HOMA-IR), liver function and inflammation (adiponectin and hs-CRP) in those with a fatty liver than those without fatty liver, which were independent from the waist circumference above 85 cm..

A significantly reduced adiponectin concentration was found in the pre-MetS and MetS groups as compared to the control group. Moreover, low concentrations of adiponectin were strongly associated with the prevalence of fatty liver which was found in the control, pre-MetS, and MetS groups (Table1). Adiponectin concentrations were significantly lower in cases with fatty liver versus those without fatty liver in all three groups (control, pre-MetS and MetS) (Table 2). The hs-CRP concentration was also significantly higher in patients with fatty liver versus those without fatty liver in all three groups (Table 2). HOMA-IR was significantly higher in pre-MetS and MetS cases with fatty liver versus those without fatty liver. HDL-C was reduced similarly with adiponectin and reduced significantly with fatty liver in all three groups. Choline esterase (ChE) displayed the same trends as adiponectin, hsCRP and RLP-TG. Also, the cases with fatty liver exhibited significantly higher concentrations of ALT, AST, γGTP and ALP than those without a fatty liver, although most of these concentrations were within the normal cut-off values.

The multiple regression analysis of cardiovascular risk markers for their association with the status of fatty liver

Table 3 shows that results of multiple regression analysis with either of adiponectin, HOMA-IR, hs-CRP, lipids and lipoproteins set as the target variable, and three variables set as explanatory variables: age, the presence of fatty liver and waist circumference (cm). The values in the Table 3 represent the standardized partial regression coefficients that correspond to partial (or adjusted) correlation coefficient between the target and each of the explanatory variables. The following findings were obtained from the table. Adiponectin and hs-CRP were more strongly associated with the presence of fatty liver than with waist circumference. In contrast, HOMA-IR, HDL-C, TG. RLP-C, RLP-TG, sd LDL-C and LDL-C were more strongly associated with waist circumference (cm) than with the presence of fatty liver. TC was associated neither with the presence of fatty liver nor with the waist circumference.

Table 3.

Multiple regression analysis of cardiovascular risk markers associated with age, fatty liver and waist circumference.

Target variables	data size	Explanatory variables (standard partial regression coefficient)
Target variables	data size	Age	Presence of fatty liver*	Waist circumference (cm)
Adiponectin	412	0.201^†	−0.301^†	−0.125
HOMA-IR	217	−0.010	0.241^†	0.405^†
Hs—CRP	332	0.179^†	0.224^†	0.143
HDL-C	810	0.096^‡	−0.200^†	−0.273^†
TG	810	−0.022	0.188^†	0.225^†
RLP−TG	420	−0.071	0.165^‡	0.296^†
RLP-C	420	−0.034	0.152^‡	0.223^†
LDL-C	810	−0.061	0.116^‡	0.132^†
sd LDL-C	332	0.006	0.097	0.319^†
TC	809	0.019	0.034	0.033

Presence or absence of fatty liver is expressed as 1 or 0, respectively.

†

P < 0.001.

‡

P < 0.01.

Multiple logistic regression analysis to identify the variables useful in diagnosing fatty liver

To explore variables which are effective in diagnosing fatty liver from among the cardiovascular makers (age, ALT ChE, HDL-C, adiponectin and HbA1c), multiple logistic regression analysis was performed by use of the backward elimination procedure. Table 4 shows the variables selected as the best regression model. The important predictors of fatty liver were ALT, HbA1c, HDL-C and adiponectin in the descending order. The replacement of HbA1c with HOMA-IR as a predictor of fatty liver showed better results. The accuracy of the prediction by the model was evaluated by ROC analysis and the area under the curve (AUC) was at a high concentration of 0.875. After removing the strongest predictor ALT (Table 5), HbA1c, adiponectin and HDL-C were proved effective in that order in diagnosing fatty liver with the model’s accuracy of AUC = 0.831.

Table 4.

The relative contribution of the variables in diagnosing the status of fatty liver by multiple logistic regression analysis.

(n = 394)
Order	Parameter	β	SE (β)	z	P
0	−8.6315	2.584
1	Age	0.0333	0.018	1.823	0.0682
2	ALT	1.9166	0.298	6.432	0.0000
3	ChE	0.1061	0.045	2.359	0.0183
4	HDL-C	−0.4194	0.131	3.201	0.0014
5	Adiponectin	−1.1323	0.393	2.882	0.0040
6	HbA1c	2.8541	0.685	4.169	0.0000

β = partial regression coefficient, z = β/SE(β), and SE(β) = standard error of β.

Table 5.

The relative contribution of the variables without ALT in diagnosing the status of fatty liver by multiple logistic regression analysis.

(n = 394)
Order	Parameter	β	SE (β)	z	P
0	−1.5053	2.106
1	Age	0.0046	0.016	0.283	0.7769
2	ChE	0.1238	0.042	2.932	0.0034
3	HDL-C	−0.4134	0.119	3.461	0.0005
4	Adiponectin	−1.5568	0.37	4.207	0.0000
5	HbA1c	3.2483	0.647	5.019	0.0000

β= partial regression coefficient, z = β/SE(β), and SE(β) = standard error of β.

Factor analysis for evaluation of commonality among the cardiovascular parameters and liver enzymes

The test results of 13 mutually related cardiovascular markers and liver enzymes were applied to the exploratory factor analysis in order to investigate the commonality among them. By use of the principal component method with varimax rotation of the factorial axes, we obtained three factors as shown in Table 6. Values in the table correspond to factor loadings, which represent the correlation coefficient between the identified common factor and each of the original variables. From the pattern of factor loadings, Factor 1 was found to be highly positively correlated with TG, RLP-C, RLP-TG and sd LDL-C, and slightly with GGT. Therefore, the factor was interpreted as reflecting serum lipid concentration and accumulation of visceral fat, and thus, coined as ‘visceral fat: VF’. Factor 2 is interpreted as representing liver injury with high correlations to AST, ALT and GGT, and thus the factor was coined as ‘liver injury: LI’. Factor 3 showed moderate positive correlations with ALT, ChE, LDL-C, sd LDL-C and HOMA-IR and moderate negative correlations with HDL-C and adiponectin. Therefore, the factor was interpreted as representing the status of fatty liver (FL). The scores of three factors were computed for each individual. The accuracy of each score to differentiate the status of fatty liver was evaluated by ROC analysis. The area under the ROC curve for Factors 1, 2 and 3 were 0.636, 0.703 and 0.822, respectively, confirming a higher specificity of Factor 3 to the status of fatty liver. It is noteworthy that, although hs-CRP was more significantly associated with the presence of fatty liver than with waist circumference by the multiple regression analysis, it was found to be only weakly correlated with Factor 3 representing the status of fatty liver.

Table 6.

Factor analysis for evaluation of commonality among the cardiovascular parameters and liver enzymes.

	Factor 1	Factor 2	Factor 3
	VF: visceral fat (waist)	LI: liver injury	FL: fatty liver
AST	0.0679	0.8457	0.0530
ALT	0.0783	0.8583	0.3486
GGT	0.2994	0.5491	0.0910
ChE	0.1862	0.1244	0.5035
TG	0.8697	0.1414	0.2722
HDL-C	−0.1384	0.0669	−0.4954
LDL-C	0.1520	0.0744	0.5831
RLP-C	0.8324	0.1015	0.2423
RLP-TG	0.9252	0.1879	0.2304
sd LDL-C	0.4845	0.1847	0.4831
Adiponectin	−0.2060	−0.1937	−0.4166
CRP	0.0796	0.1741	0.2671
HOMA-IR	0.0819	0.1658	0.5856
Contribution (%)	0.2283	0.1950	0.1567

The prevalence of low adiponectin concentration (<4 mg/L) with fatty liver in controls, pre-MetS and MetS

A reduced plasma adiponectin concentration of less than 4 mg/L has been recommended as the cut-off value for cardiovascular risk assessment in the Japanese population.⁸ Table 7 shows that the prevalence of reduced adiponectin concentrations (<4 mg/L) was 1.3% (5/382) in controls, 2.3% (7/305) in pre-MetS and 14.8% (18/122) in MetS. The prevalence of a fatty liver with adiponectin concentrations less than 4 mg/L was 86% in MetS and this was higher than in the controls (56%) and pre-MetS (46%) but not statistically significant. The cases with both fatty liver and reduced adiponectin concentration (<4 mg/L) were found to be more than 6.4-fold higher in MetS compared with controls and pre-MetS.

Table 7.

Frequency of the fatty liver in cases with serum adiponectin concentrations lower than 4 mg/L in normal, pre-MetS and MetS groups.

	Normal (A)	Pre-metabolic syndrome (B)	Metabolic syndrome (C)	P
	Normal (A)	Pre-metabolic syndrome (B)	Metabolic syndrome (C)	A vs. B	A vs. C	B vs. C
Non-fatty liver	4/333 (1.2%)	8/185 (4.3%)	3/46 (6.5%)	<0.05	<0.05	NS
Fatty liver	5/49 (10.2%)	7/120 (5.8%)	18/76 (23.7%)	NS	NS	<0.001
Ratio of fatty liver in cases with adiponectin <4 mg/L	5/9 (56%)	7/15 (46%)	18/21 (86%)	NS	NS	NS
Ratio of cases with both fatty liver and adiponectin<4 mg/L in all cases	5/382 (1.3%)	7/305 (2.3%)	18/122 (14.8%)	NS	<0.001	<0.001

Discussion

Coronary artery disease and CKD, but not chronic liver disease, have been reported to be common causes of death in patients with NAFLD.^3,21,22 Therefore, establishing a diagnostic marker for fatty liver associated with CVD is of significant clinical importance. Although certain fatty liver markers, such as ChE, liver enzymes and the markers of dysglycemia are already well established, those diagnostic markers are elevated as a consequence of liver injuries. More rigorous and predictive plasma markers associated with fatty liver and CVD are required in routine clinical practice that can be used together with ultrasonography at general hospitals. This is the first report describing the comparative study among the plasma cardiovascular risk markers, the presence of fatty liver, waist circumference above 85 cm and their associations with liver injury markers in the three different groups of Japanese men.

We have found a significantly higher prevalence of fatty liver in MetS as compared to pre-MetS or controls in our hospital. Therefore, the primary aim of this study was to evaluate the diagnostic markers of fatty liver which are also recognized as cardiovascular risk markers such as adiponectin, hs-CRP, HOMA-IR, HDL-C, LDL-C and RLPs. Another aim of this study was to clarify whether fatty liver is always associated with MetS or plays an important role in predicting the risk of coronary heart disease (CHD) in much the same way as increased visceral fat has been proposed to predict the risk of CHD, especially by evaluating the plasma concentrations of adiponectin.

The anthropometric parameters, such as waist circumference and BMI, differed in the groups with and without fatty liver (Table 2). Our data also show that fatty liver was associated with reduced adiponectin,^23,24 increased hs-CRP^13,25 and a high degree of insulin resistance, expressed as a higher HOMA-IR index¹³ as reported previously. Furthermore, increased RLP-C was observed in the fatty liver group with the higher concentration of insulin resistance.^26–28 Other characteristics which were evaluated hepatic enzymes, e.g. ALT and other enzymes associated with liver injuries, and those were mildly to moderately increased in patients with a fatty liver, suggesting that these enzymes were the consequences of liver injury pathogenesis, but not predictors of CVD associated with fatty liver. HOMA-IR has been known to be a marker of diabetes and known as a risk marker for CVD.

The adiponectin concentrations exhibited significant reduction in all three groups with fatty liver similar to HDL-C, regardless of the presence of MetS. Adiponectin secretion has been reported to be suppressed by TNF-α, IL-6, etc. or when adipocytes enlarged with the accumulation of lipids.²⁹ Therefore, when the waist circumference increases, the enlarged adipocytes in MetS and pre-MetS have been said to reduce adiponectin secretion compared to normal adipocytes in cases with normal waist circumference. However, we did not find the decreased plasma adiponectin concentrations in MetS and pre-MetS cases without fatty liver who had larger waist circumference above 85 cm. Therefore, the decreased plasma adiponectin concentrations in cases with fatty liver may be the consequence of increased hepatic uptake of adiponectin from the blood in order to protect against damages due to fatty liver. Whereas, in the MetS and pre-MetS cases without fatty liver, the hepatic uptake of adiponectin from the blood is not required, and plasma adiponectin concentrations remained normal concentrations. In those cases, adiponectin in enlarged adipocytes may be synthesized and secreted normally as in normal adipocytes. Therefore, we can say that decreased plasma adiponectin concentrations reflect the strong association with the presence of fatty liver, but independent from the visceral fat volume or waist circumference above 85 cm.

CRP has been measured to detect the chronic inflammatory state and taking into account the fact that CRP plays an important role in the progression of NAFLD and atherosclerosis.^30,31 A significant increase in hs-CRP concentration with fatty liver was observed in MetS, pre-MetS and normal control cases compared to the cases without fatty liver in our study. Some authors have reported higher hs-CRP concentrations in patients with fatty liver and even proposed it to be a predictor of NAFLD.^13,25 On the other hand, Musso et al.³² did not find any difference in hs-CRP concentration upon conducting an intragroup analysis of insulin resistant patients, comparing those with and without fatty liver. In our findings, there was an increase in hs-CRP among the normal controls, pre-MetS and MetS cases, all linked to the presence of fatty liver regardless of the presence of MetS. CRP may thus play a role in the CHD associated with fatty liver. Ridker et al.³³ reported in the JUPITER trial that the statin significantly reduced cardiovascular events in cases with higher hs-CRP concentrations regardless of LDL-C concentrations. These results may be associated with the attenuation of fatty liver more rather than the reduction of LDL-C by statin treatment, although they did not assess the relationship between the CHD risk reduction and the presence of fatty liver in JUPITER trial.

The relationship between MetS and plasma lipoprotein concentrations such as TG and HDL-C has been well established.¹⁹ RLPs are also reported to increase significantly in MetS.²⁰

Although both RLP-C and RLP-TG increased significantly in MetS compared to normal controls and pre-MetS, the relationship between the cases with and without fatty liver indicated the different trends from adiponectin and hsCRP. RLP-TG was significantly higher in controls and pre-MS with fatty liver compared to the non-fatty liver cases, but not different in MetS. However, RLP-C was not different between the cases with and without fatty liver in all the three groups, although RLP-C significantly increased in MetS compared with pre-MetS and controls. These results indicate that formation of RLP-C is controlled by a different metabolic pathway from that of RLP-TG. RLP-C formation is known to be strongly associated with cholesteryl ester transfer protein activity in plasma,^34–36 while RLP-TG formation is known to be associated with LPL activity which is partly activated by adiponectin.^37–40 RLP-TG was most strongly associated with waist circumference among the variables studied, indicating a specific marker of MetS by factor analysis (Table 6). The ratio of RLP-TG/RLP-C is known to reflect the particle size of RLPs,^41–43 and RLP particles in fatty liver were shown to be larger particle sizes than those without fatty liver in controls and pre-MetS, but not different in MetS which was significantly larger in both cases (Table 2).

Another significant lipoproteins associated with fatty liver were HDL-C. HDL-C decrease was very similar to adiponectin decrease in cases with fatty liver compared to those without fatty liver in all three groups. Furthermore, HDL-C was also linked to variables associated with liver injury markers and adiponectin as determined by factor analysis (Table 6). Therefore, adiponectin and HDL-C were significantly correlated with liver injury markers which may also be associated with the risk of CVD. The decrease in HDL-C in fatty liver may be due to a similar mechanism to that which lowers adiponectin. Both variables may reduce the plasma concentrations in order to protect against liver injury including oxidative stress damages in fatty liver.

Table 3 shows that adiponectin and hs-CRP were stronger diagnostic markers of fatty liver than waist circumference by multiple regression analyses. Since waist circumference and BMI are highly correlated, we have used waist circumference instead of BMI for this analysis followed by the definition of MetS. HOMA-IR, HDL-C, TG, RLP-C, RLP-TG and sd LDL-C were stronger markers of waist circumference (cm) than the presence of fatty liver. These results suggest that adiponectin and hs-CRP were more strongly associated with the presence of fatty liver, independent of visceral fat volume, regardless of the presence of MetS. However, most lipoproteins were more strongly associated with the visceral fat volume than the presence of fatty liver.

Furthermore, the multiple logistic regression analysis (Table 4) revealed that the important predictors of fatty liver were ALT, HbA1c, HDL-C and adiponectin. The replacement of HbA1c with HOMA-IR in the model resulted in similar manner in the accuracy for diagnosing fatty liver in the same category of insulin resistance. This priority of HbA1c over HOMA-IR in the diagnosis is of clinical importance because HbA1c measurement is more commonly ordered than HOMA-IR at hospital laboratories together with liver enzymes. The addition of adiponectin measurement to the model led to better prediction of fatty liver with higher accuracy. After removing the strongest predictor of ALT (Table 5), HbA1c, adiponectin and HDL-C were proved to be more effective in diagnosing fatty liver. On the other hand, another logistic analysis was performed to explore variables which are effective in diagnosing MetS with waist circumference above 85 cm. RLP-TG and fasting glucose were selected as significant (data not shown). However, adiponectin was not included as a significant variable for predicting MetS. These results indicated the same with the previous study reported by Speliotes et al.¹⁰ at Framingham Study that fatty liver is independent from the visceral adipose tissue when predicted by plasma adiponectin concentrations.

As an additional investigation, we obtained the three factors by exploratory factor analysis as shown in Table 6. Three factors were interpreted as reflecting or representing visceral fat (Factor 1), liver injury (Factor 2) and fatty liver (Factor 3), respectively. Factor 3 showed moderate positive correlation with ALT, ChE, LDL-C, sd LDL-C and HOMA-IR and also moderate negative correlations with HDL-C, and adiponectin. Therefore, Factor 3 was interpreted as representing the status of fatty liver. As a whole, we interpreted the findings as follows: ALT represents both liver injury and fatty liver factors; sd LDL-C represents both visceral fat and fatty liver. Factors such as ChE, LDL-C, HDL-C, adiponectin and HOMA-IR primarily represent fatty liver factor, while TG, RLP-C, RLP-TG primarily represent visceral fat factor. Although hs-CRP was shown to be more significantly associated with the presence of fatty liver than with waist circumference in Table 3, only slight correlations were found with Factor 3 by the factor analysis. The individual scores of Factor 3 (FL-score) were computed and the diagnostic utility of the FL-s core in diagnosing the presence of fatty liver was examined. The AUC representing the accuracy of the diagnosis was 0.822 (data not shown). The use of Factor 3 markers which are most commonly ordered by routine clinical practice at hospital laboratories could bring more accurate detection of fatty liver when adiponectin is added.

The prevalence of fatty liver cases with plasma adiponectin concentration less than 4 mg/L, which has been recommended as the adiponectin cut-off value for MetS or cardiovascular risk assessment in the Japanese population^8,31 was 86% in MetS, with significantly lower prevalence in the controls and pre-MetS cases, as shown in Table 7. These results suggest that the co-presence of fatty liver and low adiponectin concentration in MetS is strongly associated with CHD risk.^44–47 High prevalence of CHD in MetS thus appears to be closely associated with the presence of fatty liver.^48,49 Recently, Ai et al.⁵⁰ reported that plasma hypoadiponectin is an independent risk factor for CHD in men in the Framingham Offspring Study. Those results may be associated with the presence of fatty liver in cases with low adiponectin concentrations, although they did not assess liver fat status in this study. Further studies need to be done by multidetector computed tomography examinations of the abdomen which allow for a more precise quantification compared with indirect measures of fatty liver by ultrasonography.

In conclusion, we have shown that adiponectin was found to be a more specific diagnostic marker of fatty liver than waist circumference, regardless of the presence of MetS. Adiponectin was also linked live enzyme, HbA1c and HDL-C, indicating the association between the liver injury pathogenesis and the risk of CVD. RLPs, especially RLP-TG, were found to be the most specific marker of waist circumference, namely visceral fat volume, but not of fatty liver. By adding the measurement of plasma adiponectin concentrations as Fatty Liver- score, we believe that it could predict the CHD risk with higher accuracy in patients with fatty liver in addition to standard risk factors such as HDL-C.

Footnotes

Acknowledgements

The authors deeply thank Professor Kiyoshi Ichihara of Yamaguchi University Graduate School of Medicine for his valuable discussions on the statistical analysis in this study.

Declaration of conflicting interests

The authors declare no conflict of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Ethical approval

The study was conducted following informed consent and was approved by the Ethical Committees of Hidaka Hospital and of Gunma University School of Medicine.

Guarantor

KN.

Contributorship

All authors confirmed they have contributed to the intellectual content of this paper and have met the following three requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content and (c) final approval of the published article.

References

Eguchi

Hyogo

Ono

. Prevalence and associated metabolic factors of nonalcoholic fatty liver disease in the general population from 2009 to 2010 in Japan: a multicenter large retrospective study. J Gastroenterol 2012; 47: 586–595.

Targher

Chonchol

Pichiri

. Risk of cardiovascular disease and chronic kidney disease in diabetic patients with non-alcoholic fatty liver disease: just a coincidence? J Endocrinol Invest 2011; 34: 544–551.

Grundy

Cleeman

Daniels

. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005; 112: 2735–2752.

Byrne

Olufadi

Bruce

. Metabolic disturbances in non-alcoholic fatty liver disease. Clin Sci (Lond) 2009; 116: 539–564.

Okamoto

Kihara

Funahashi

. Adiponectin: a key adipocytokine in metabolic syndrome. Clin Sci (Lond) 2006; 110: 267–278.

Matsuzawa

Funahashi

Kihara

. Adiponectin and metabolic syndrome. Arterioscler Thromb Vasc Biol 2004; 24: 29–33.

Bremer

Devaraj

Afify

. Adipose tissue dysregulation in patients with metabolic syndrome. J Clin Endocrinol Metab 2011; 96: E1782–E1788.

Ryo

Nakamura

Kihara

. Adiponectin as a biomarker of the metabolic syndrome. Circ J 2004; 68: 975–981.

Açikel

Sunay

Koplay

. Evaluation of ultrasonographic fatty liver and severity of coronary atherosclerosis, and obesity in patients undergoing coronary angiography. Anadolu Kardiyol Derg 2009; 9: 273–279.

10.

Speliotes

Massaro

Hoffmann

. Fatty liver is associated with dyslipidemia and dysglycemia independent of visceral fat: the Framingham Heart Study. Hepatology 2010; 5: 1979–1987.

11.

Tilg

Moschen

. Insulin resistance, inflammation, and non-alcoholic fatty liver disease. Trends Endocrinol Metab 2008; 19: 371–379.

12.

Vanni

Bugianesi

Kotronen

. From the metabolic syndrome to NAFLD or vice versa? Dig Liver Dis 2010; 42: 320–323.

13.

Park

Kim

Yun

. Insulin resistance and C-reactive protein as independent risk factors for non-alcoholic fatty liver disease in non-obese Asian men. J Gastroenterol Hepatol 2004; 19: 694–698.

14.

Gómez Rosso

Benítez

Fornari

. Alterations in cell adhesion molecules and other biomarkers of cardiovascular disease in patients with metabolic syndrome. Atherosclerosis 2008; 199: 415–423.

15.

Shin

Kim

. Remnant lipoprotein particles induce apoptosis in endothelial cells by NAD (P)H oxidase-mediated production of superoxide and cytokines via lectin-like oxidized low-density lipoprotein receptor-1 activation: prevention by cilostazol. Circulation 2004; 109: 1022–1028.

16.

Nakajima

Nakano

Tanaka

. The oxidative modification hypothesis of atherosclerosis: the comparison of atherogenic effects on oxidized LDL and remnant lipoproteins in plasma. Clin Chim Acta 2006; 367: 36–47.

17.

Adiels

Taskinen

Packard

. Overproduction of large VLDL particles is driven by increased liver fat content in man. Diabetologia 2006; 49: 755–765.

18.

Watts

Farvid

. Adipocytokines and VLDLmetabolism: independent regulatory effects of adiponectin, insulin resistance, and fatcompartments on VLDL apolipoprotein B-100 kinetics? Diabetes 2005; 54: 795–802.

19.

Teramoto

Sasaki

Ueshima

. Metabolic syndrome. J Atheroscler Thromb 2008; 15: 1–5.

20.

Nakano

Tokita

Nagamine

. Measurement of serum remnant-like lipoprotein particle-triglyceride (RLP-TG) and RLP-TG/ total TG ratio using highly sensitive triglyceride assay reagent. Clin Chim Acta 2011; 412: 71–78.

21.

Kamioka

Suzuki

Yamada

. High sensitivity C-reactive protein predicts nonresponders and cardiac deaths in severe heart failure patients after CRT implantation. Int Heart J 2012; 53: 306–312.

22.

Nseir

Shalata

Marmor

. Mechanisms linking nonalcoholic fatty liver disease with coronary artery disease. Dig Dis Sci 2011; 56: 3439–3449.

23.

Targher

Bertolini

Zenari

. Hypoadiponectinemia is closely associated with nonalcoholic hepatic steatosis in obese subjects. Diabetes Care 2004; 27: 2085–2086.

24.

Targher

Bertolini

Rodella

. Associations between plasma adiponectin concentrations and liver histology in patients with nonalcoholic fatty liver disease. Clin Endocrinol (Oxf) 2006; 64: 679–683.

25.

Kogiso

Moriyoshi

Shimizu

. High-sensitivity C-reactive protein as a serum predictor of nonalcoholic fatty liver disease based on the Akaike Information Criterion scoring system in the general Japanese population. J Gastroenterol 2009; 44: 313–321.

26.

Watanabe

Taniguchi

Taketoh

. Elevated remnant-like lipoprotein particles in impaired glucose tolerance and type 2 diabetic patients. Diabetes Care 1999; 22: 152–156.

27.

Abbasi

McLaughlin

Lamendola

. Fasting remnant lipoprotein cholesterol and triglyceride concentrations are elevated in nondiabetic, insulin-resistant, female volunteers. J Clin Endocrinol Metab 1999; 84: 3903–3906.

28.

Kotani

Asahara-Satoh

Kato

. Remnant-like particle cholesterol and serum amyloid A-low-density lipoprotein levels in obese subjects with metabolic syndrome. J Clin Lipidol 2011; 5: 395–400.

29.

Stern

Osher

Greenman

. Hypoadiponectinemia as a marker of adipocyte dysfunction – Part I: the biology of adiponectin. J Cardiometab Syndr 2007; 2: 174–182.

30.

Lucero

Zago

López

. Pro-inflammatory and atherogenic circulating factors in non-alcoholic fatty liver disease associated to metabolic syndrome. Clin Chim Acta 2011; 412: 143–147.

31.

Hiuge-Shimizu

Kishida

Funahashi

. Coexistence of visceral fat and multiple risk factor accumulations is strongly associated with coronary artery disease in Japanese (The VACATION-J Study). J Atheroscler Thromb 2012; 19: 657–663.

32.

Musso

Gambino

. Should nonalcoholic fatty liver disease be included in the definition of metabolic syndrome? Across-sectional comparison with Adult Treatment Panel III criteria in nonobese nondiabetic subjects. Diabetes Care 2008; 31: 562–568.

33.

Ridker

Fonseca

Genest

. Baseline characteristics of participants in the JUPITER trial, a randomized placebo-controlled primary prevention trial of statin therapy among individuals with low low-density lipoprotein cholesterol and elevated high-sensitivity C-reactive protein. Am J Cardiol 2007; 100: 1659–1664.

34.

De Grooth

Klerkx

Stroes

. A review of CETP and its relation to atherosclerosis. J Lipid Res 2004; 45: 1967–1974.

35.

Okamoto

Miyai

Sasase

. Cholesteryl ester transfer protein promotes the formation of cholesterol-rich remnant like lipoprotein particles in human plasma. Clin Chim Acta 2007; 375: 92–98.

36.

Tanaka

Shimokado

. A deficiency of cholesteryl ester transfer protein whose serum remnant-like particle-triglyceride significantly increased, but serum remnant-like particle-cholesterol did not after an oral fat load. Ann Clin Biochem 2009; 46: 457–463.

37.

De Vries

Wolffenbuttel

Sluiter

. Post-heparin plasma lipoprotein lipase, but not hepatic lipase activity, is related to plasma adiponectin in type 2 diabetic patients and healthy subjects. Clin Lab 2005; 51: 403–409.

38.

Kobayashi

Kusunoki

Murase

. Relationship of lipoprotein lipase and hepatic triacylglycerol lipase activity to serum adiponectin levels in Japanese hyperlipidemic men. Horm Metab Res 2005; 37: 505–509.

39.

Qiao

Zou

van der Westhuyzen

. Adiponectin reduces plasma triglyceride by increasing VLDL triglyceride catabolism. Diabetes 2008; 57: 1824–1833.

40.

Ganguly

Schram

Fang

. Adiponectin increases LPL activity via RhoA/ROCK-mediated actin remodelling in adult rat cardiomyocytes. Endocrinology 2011; 152: 247–254.

41.

Okazaki

Usui

Tada

. Relation between RLP-triglyceride to RLP-cholesterol ratio and particle size distribution in RLP-cholesterol profiles by HPLC. Clin Chim Acta 2000; 296: 135–149.

42.

Nakajima

Nakano

Tokita

. Postprandial lipoprotein metabolism: VLDL vs chylomicrons. Clin Chim Acta 2011; 412: 1306–1318.

43.

Adiels

Taskinen

Packard

. Overproduction of large VLDL particles is driven by increased liver fat content in man. Diabetologia 2006; 49: 755–765.

44.

Alper

Hasdemir

Sahin

. The relationship between nonalcoholic fatty liver disease and the severity of coronary artery disease in patients with metabolic syndrome. Turk Kardiyol Dern Ars 2008; 36: 376–381.

45.

Thiruvagounder

Khan

Sheriff

. Non-alcoholic fatty liver disease (NAFLD) - Is it an emerging risk factor for coronary artery disease?: preliminary study in a local Indian population. Sultan Qaboos Univ Med J 2010; 10: 221–226.

46.

Targher

Bertolini

Padovani

. Prevalence of nonalcoholic fatty liver disease and its association with cardiovascular disease among type 2 diabetic patients. Diabetes Care 2007; 30: 1212–1218.

47.

Assy

Djibre

Farah

. Presence of coronary plaques in patients with nonalcoholic fatty liver disease. Radiology 2010; 254: 393–400.

48.

Ahmed

Barakat

Almobarak

. Nonalcoholic fatty liver disease and cardiovascular disease: has the time come for cardiologists to be hepatologists? J Obes 2012; 2012: 483135–483135.

49.

Nestel

Mensink

. Perspective: nonalcoholic fatty liver disease and cardiovascular risk. Curr Opin Lipidol 2013; 24: 1–3.

50.

Otokozawa

Asztalos

. Adiponectin: an independent risk factor for coronary heart disease in men in the Framingham offspring study. Atherosclerosis 2011; 217: 543–548.

Plasma adiponectin is a more specific marker of fatty liver than a marker of metabolic syndrome in Japanese men

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Materials and methods

Subjects

Biochemical analysis

Statistical analysis

Results

The clinical and biochemical characteristics in controls, pre-MetS and MetS cases

The comparison of diagnostic parameters among the cases with or without fatty liver in controls, pre-MetS and MetS cases

The multiple regression analysis of cardiovascular risk markers for their association with the status of fatty liver

Multiple logistic regression analysis to identify the variables useful in diagnosing fatty liver

Factor analysis for evaluation of commonality among the cardiovascular parameters and liver enzymes

The prevalence of low adiponectin concentration (<4 mg/L) with fatty liver in controls, pre-MetS and MetS

Discussion

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

Ethical approval

Guarantor

Contributorship

References