Obesity Research & Clinical Practice (2010) 4, e217—e224
ORIGINAL ARTICLE
Metabolic syndrome predicts the incidence of hepatic steatosis in Koreans Kayoung Lee ∗ Department of Family Medicine, Busan Paik Hospital, Inje University College of Medicine, 633-165 Kaegum-dong, Busan Jin-Gu, Busan 614-735, South Korea Received 5 November 2009 ; received in revised form 30 January 2010; accepted 2 February 2010
KEYWORDS Metabolic syndrome; Hepatic steatosis; Incidence
∗
Summary Problem: It has not been well elucidated whether the development of metabolic syndrome and its components predicts the incidence of hepatic steatosis. Methods: A cohort of 1705 apparently healthy Korean adults (954 men and 751 women, 43.6 ± 8.5 years old) without ultrasonographically defined hepatic steatosis and with normal serum gamma-glutamyl-transpeptidase and alanine aminotransferase was followed from 2004 to 2007. Metabolic syndrome was defined as the presence of at least three of the following components: obesity (BMI ≥ 25.0 kg/m2 ), high blood pressure, elevated levels of triglycerides and fasting glucose, and low level of high-density lipoprotein cholesterol. The outcome was ultrasonographically diagnosed hepatic steatosis. The analyses were conducted using the Cox proportional hazards model and time-dependent Cox model. Results: A total of 226 individuals developed hepatic steatosis during 3716 person—years of follow-up. The presence of one to two metabolic syndrome components at baseline significantly predicted the development of hepatic steatosis. Metabolic syndrome itself, having ≥1 metabolic syndrome components, and maintenance of metabolic syndrome during follow-up significantly increased the risk (hazard ratio 2.0—4.1 for men, 3.4—10.8 for women) after adjustment for the followup period, age and BMI at baseline or updated during follow-up. Occurrence of obesity, high triglycerides or high fasting glucose during follow-up significantly predicted the development of hepatic steatosis, even after adjustment for metabolic syndrome components at baseline. Conclusions: The presence at baseline and the development of metabolic syndrome during follow-up were risk factors for ultrasonographically detected hepatic steatosis. © 2010 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
Tel.: +82 51 890 6229; fax: +82 51 894 7554. E-mail addresses: [email protected], [email protected].
1871-403X/$ — see front matter © 2010 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2010.02.004
e218
Introduction Hepatic steatosis is a common hepatic disorder characterized by fat accumulation in the liver seen in individuals who excessively drink alcohol (alcohol-related fatty liver) or in those without history of relevant alcohol consumption (nonalcoholic fatty liver disease, NAFLD) [1,2]. NAFLD is frequently associated with conditions manifested by insulin resistance, such as obesity, hypertension, dyslipidemia, type 2 diabetes, and metabolic syndrome (MetS) [3—5]. There is evidence that NAFLD is related to higher mortality caused by liver-related complications [6,7] and to an elevated risk of cardiovascular disease [8]. While several cross-sectional studies have demonstrated the co-existence of metabolic syndrome and NAFLD [9—11], only one cohort study addressed the association between the presence of MetS at baseline and the future development of NAFLD [12]. However, the use of self-reported information for alcohol consumed may be prone to error and may bias the differentiation between NAFLD and alcohol-related liver disease. Studies have reported that self-reports appear to underestimate alcohol consumption in a sizable proportion of patients in clinical studies [13]. In contrast, serum gamma-glutamyl-transpeptidase (GGT) and alanine aminotransferase (ALT) has been shown to be sensitive markers of liver disease related with alcohol consumption [14,15]. Therefore, the clarification of relationship between MetS and incidence of hepatic steatosis in a prospective study among participants with normal range of ALT and GGT is helpful to extend knowledge about the temporal relationships between the two conditions among individuals after excluding alcohol-related liver damage. This study was carried out to evaluate the role of MetS and its components, at baseline and during follow-up, as risk factors for the development of ultrasonographically diagnosed hepatic steatosis among Korean men and women with normal range of ALT and GGT.
Methods Study subjects This study used data generated from 7553 Korean adults (4336 men and 3217 women, aged 20 years or older) who visited the Center of Health Promotion at Inje University, Busan Paik Hospital in 2004 to have health examinations. Subjects with evidence of excessive alcohol intake (≥20 g/day), abnormal
K. Lee level of GGT (>50 IU/L) and ALT (> 45 IU/L), positive seromarkers for hepatitis B or C, biliary disease, liver cirrhosis, or malignant disease were excluded through laboratory tests and ultrasonography at baseline. Alcohol intake was assessed using two open questions; ‘‘How often did you have a drink containing alcohol per week in the past 6 months?’’ and ‘‘How many glasses did you have on a typical day when you were drinking in the past 6 months? In the second question, one glass of alcoholic beverage was assumed to contain 10 g of alcohol. From these two questions, we calculated an average daily intake of alcohol. A total of 5986 individuals were eligible for the study at baseline, and of these, 1705 apparently healthy individuals (954 men and 751 women, 43.6 ± 8.5 years old) voluntarily received at least two evaluations and without specific regular interval but at least after one year from baseline examination, including abdominal ultrasonography, biochemical tests for liver and metabolic function, and measurements of body mass index (BMI), from 2004 until December of 2007. Compared to those excluded in the analyses (N = 4281), the individuals finally included in analyses were younger, and less likely to be women and fulfilled with the criteria of MetS components.
Measurements Health examinations performed at baseline and during follow-up were abdominal ultrasonography, biochemical tests for liver and metabolic function, blood pressure (BP) measurement, and anthropometric measurements. Hepatic steatosis was diagnosed using abdominal B-mode ultrasonography (LOGIQ 7 with a 4 MHz transducer; GE Healthcare, Wauwatosa, WI, USA). Ultrasonography was performed by experienced radiologists who did not have access to subjects’ clinical and laboratory test findings. Hepatic steatosis was originally classified from mild to severe levels but defined in current analyses when at least the criteria of mild hepatic steatosis were satisfied: slight increase in hepatic echogenicity and differences between hepatic and renal echogenicity, and relative preservation of echoes from the walls of the portal vein [16]. Body weight and height were measured in subjects clothed in a light gown without shoes; these measurements were used to calculate the BMI (kg/m2 ). A BMI of 25 kg/m2 or greater, which has been proposed for Asians as indicative of obesity, was used to define the presence of obesity [17]. Blood pressure measurements were performed with a standard manual sphygmomanometer while participants were in a sitting position. Antecubital venous blood samples were
Metabolic syndrome predicts NAFLD taken from all subjects after a 12-h overnight fast. The levels of high-density lipoprotein cholesterol (HDL-C), triglycerides (TGs), fasting glucose (FG), alanine aminotransferase (ALT), and gammaglutamyl-transpeptidase (GGT) were measured using a biochemistry autoanalyzer (Toshiba 200FR; Toshiba, Tokyo, Japan). The following criteria were used to define abnormalities of the MetS components: BP ≥ 130/85 mmHg; fasting glucose (FG) ≥ 100 mg/dL; high-density lipoprotein cholesterol (HDL-C) < 40 mg/dL for men or < 50 mg/dL for women; and TG ≥ 150 mg/dL [18]. Individuals with at least three abnormal MetS components, among the four components mentioned above and overall obesity (abdominal obesity was not used, because waist circumference was not assessed). The ethics committee of Inje University, Busan Paik Hospital approved this study.
Statistical analyses The 2 -test and t-test were used to assess the association between the baseline characteristics and the development of hepatic steatosis. Multivariate Cox proportional hazards modeling of the development of hepatic steatosis was performed to investigate relationships with each MetS component at baseline, after adjustment for follow-up duration, other MetS components, sex, and age at baseline. The combinations of the presence of abnormal MetS components and MetS at baseline and at follow-up assessments for each subject were categorized into four groups: no MetS at both assessments, change from non-MetS to MetS, change from MetS to no MetS, and having MetS at both tests. The follow-up assessment was defined as the assessment performed at a follow-up visit when hepatic steatosis was diagnosed for incident cases or the final assessment for non-cases. The Cox proportional hazards model and time-dependent Cox regression model were used separately in men and women to assess the risk for developing hepatic steatosis across the four categories of MetS, after adjustment for age, and BMI at baseline. In the time-dependent Cox regression analysis, the MetS categories and the change in covariates over follow-up time were modeled as time-dependent variables. The follow-up duration was calculated as the interval (as years) from the first visit until the time of hepatic steatosis development or until the final assessment, for each individual. Cox proportional hazard model was also used to assess the risk for incident hepatic steatosis with each MetS component category, after adjustment for age, sex, and various baseline-measurements (BMI, systolic BP, TG, HDL-C, and FG). Statistical significance
e219 was indicated when P < 0.05. All statistical analyses were performed using SPSS 17.0 (Release 17.0.0 (23rd, Apr. 2008); SPSS Inc., Chicago, IL, USA).
Results Longitudinal change of metabolic syndrome and hepatic steatosis Of the 1705 subjects without hepatic steatosis at baseline, the prevalence of MetS changed from 7.4% at baseline to 7.2% on follow-up: obesity, from 22.6% to 24.3%; high BP, from 19.8% to 18.0%; high TG, from 11.7% to 14.6%; low HDL-C, from 21.9% to 11.6%; and high FG, from 13.0% to 14.8%. During the follow-up, 88.8% of subjects remained free of MetS, while 3.4% of subjects (4.7% of men, 1.7% of women) continued to have MetS. For the 4.0% of subjects (4.6% of men, 3.3% of women) who had MetS at study entry, the MetS criteria were not more fulfilled during follow-up. In contrast, 3.8% of subjects (4.8% of men, 2.4% of women) without MetS at baseline developed incident MetS during the followup time. A total of 226 individuals (167 men and 59 women) developed hepatic steatosis during 3716 (2074 for men and 1642 for women) person—years of follow-up.
Risk for the development of hepatic steatosis by metabolic syndrome component at baseline Compared with the reference group, subjects with high BMI, high TG, or high FG at baseline were more likely to develop hepatic steatosis, after adjusting for the presence of other MetS components, age, and sex. The risk of developing hepatic steatosis was almost 6-fold higher in subjects who had MetS at baseline and was 2.4-fold higher even for those with one or two components at baseline (Table 1).
Risk for the development of hepatic steatosis by metabolic syndrome category during the follow-up time Table 2 shows that the risk for developing hepatic steatosis significantly increased with the development of MetS. Compared with those without MetS during follow-up, the presence of MetS at the follow-up assessment predicted the development of hepatic steatosis, after adjustment for age and BMI at baseline. The Cox regression models including the four categories of MetS, changes in BMI and other covariates over time as time-dependent
e220
Table 1
The comparison of baseline characteristics by category of metabolic syndrome. Overall, N = 1705
FL (−), N = 1479
FL (+), N = 226
Mean ± standard deviation, N (%) Age and sex-adjusted
Multivariate† — —
43.6 ± 8.5 954 (56.0)
43.3 ± 8.5 787 (53.2)
46.2 ± 8.3 167 (73.9)
— —
385 (22.6) 338 (19.8) 199 (11.7) 373 (21.9) 222 (13.0)
283 (19.1) 280 (18.9) 137 (9.3) 302 (20.4) 168 (11.4)
102 (45.1) 58 (25.7) 62 (27.4) 71 (31.4) 54 (23.9)
2.74 1.14 2.67 1.76 1.74
Number of MetS components at baseline*2 0 743 (43.6) 1—2 835 (49.0) 3—5 127 (7.4)
697 (47.1) 705 (47.7) 77 (5.2)
46 (20.4) 130 (57.5) 50 (22.1)
1.0 2.37 (1.69—3.33) 5.91 (3.93—8.89)
Age (y) Male*1
*1
MetS component at baseline BMI ≥ 25 kg/m2*2 BP ≥ 130/85 mmHg*2 TG ≥ 150 mg/dL*2 HDL-C < 40 (M) 50 (F) mg/dL*2 FG ≥ 100 mg/dL*2
HR (95% CI)
(2.11—3.57) (0.84—1.56) (1.99—3.58) (1.32—2.33) (1.26—2.40)
2.46 0.99 2.10 1.23 1.61
(1.88—3.22) (0.72—1.34) (1.52—2.89) (0.91—1.68) (1.16—2.22)
— —
Abbreviations: MetS, metabolic syndrome; BMI, body mass index; BP, blood pressure; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; FG, fasting glucose. * P < 0.05 using 1 t-test or 2 chi-square test. † Using Cox proportional hazard models after adjustment for age, sex, and other MetS components at baseline.
K. Lee
Association of metabolic syndrome (MetS) categories at baseline and follow-up with the development of hepatic steatosis. Person—years
Incident case (N, %)
Age-adjusted HR (95% CI)
Multivariate HR (95% CI)*
HR adjusted with time-dependent variables (95% CI)†
Male (N = 954) Both non-MetS (N = 819) From MetS to non-MetS (N = 44) From non—MetS to MetS (N = 46) Both MetS (N = 45)
1801 98 92 83
115 (14.0) 15 (34.1) 16 (34.8) 21 (46.7)
1.00 2.38 (1.39—4.08) 2.76 (1.64—4.66) 4.17 (2.61—6.64)
1.00 1.60 (0.91—2.81) 1.96 (1.13—3.38) 2.36 (1.40—4.01)
1.00 2.82 (1.64—4.85) 1.84 (1.01—3.35) 4.11 (2.58—6.54)
Female (N = 751) Both non-MetS (N = 695) From MetS to non-MetS (N = 25) From non-MetS to MetS (N = 18) Both MetS (N = 13)
1527 53 35 27
39 (5.6) 6 (24.0) 6 (33.3) 8 (61.5)
1.00 3.40 (1.41—8.22) 5.73 (2.37—13.86) 7.05 (2.98—16.71)
1.00 2.22 (0.90—5.43) 4.26 (1.77—10.26) 3.42 (1.39—8.36)
1.00 3.48 (1.43—8.43) 5.00 (2.06—12.16) 5.87 (2.35—14.69)
1.0 3.97 (2.52—6.24)
1.0 3.19 (1.98—5.13)
1.0 3.63 (2.31—5.72)
1.0 11.28 (4.40—28.89)
1.0 8.40 (3.16—22.33)
1.0 10.82 (4.24—27.62)
Male without any one MetS component at baseline (N = 865) No MetS component on f/u (N = 373) 855 MetS component≥ 1 on f/u (N = 492) 1038 Female without any one MetS component at baseline (N = 713) No MetS component on f/u (N = 438) 958 MetS component ≥ 1 on f/u (N = 275) 604
23(6.2) 108(22.0) 5 (1.1) 40 (14.5)
Metabolic syndrome predicts NAFLD
Table 2
*
Using Cox proportional hazard models after adjustment for age, follow-up duration, and BMI at baseline. Using extended Cox proportional hazard models with MetS category as a time-dependent variable and adjustment for age, follow-up duration, and change of BMI during the follow-up time as time-dependent variables. †
e221
Units are hazard ratio (95% confidence interval) for developing hepatic steatosis with each MetS component category using Cox proportional hazard models after controlling for age, sex, and baseline-measurement values (BMI, systolic blood pressure, fasting glucose, triacylglycerides, and high-density lipoprotein cholesterol).
1.0 1.41 (0.79—2.51) 1.78 (1.18—2.70) 1.90 (1.18—3.07) 80.4 4.8 6.6 8.2 1.0 0.52 (0.33—0.81) 1.75 (0.88—3.48) 0.66 (0.41—1.06) 75.9 13.1 2.1 8.9 1.0 1.53 (0.91—2.55) 2.08 (1.41—3.05) 2.20 (1.35—3.60) 80.2 5.2 8.1 6.5 1.0 0.73 (0.44—1.20) 1.11 (0.72—1.73) 0.96 (0.59—1.56) 74.2 7.8 6.0 12.0 1.0 0.78 (0.33—1.87) 2.77 (1.71—4.47) 1.92 (1.15—3.19)
HR (95% CI) % HR (95% CI) % HR (95% CI) % HR (95% CI)
HR (95% CI) % %
72.3 3.3 5.1 19.2 Category Both non-MetS From MetS to non-MetS From non-MetS to MetS Both MetS
FG ≥ 100 mg/dL HDL-C < 40 (M) 50 (F) mg/dL TG ≥ 150 mg/dL BP ≥ 130/85 mmHg
This longitudinal study among individuals with normal level of ALT and GGT at baseline confers some important findings on the relationship between MetS and hepatic steatosis. First, the longitudinal change in MetS itself was a significant risk factor of the incidence of hepatic steatosis. Compared with the reference group, the incidence of hepatic steatosis was 1.8—4.1-fold higher and 3.4—5.9-fold higher for men and women, respectively, who had MetS on the follow-up assessment, regardless of baseline MetS status, baseline BMI or BMI change over time. Even those who initially did not have any MetS components but had one MetS component on follow-up were at higher risk for future development of hepatic steatosis, regardless of baseline BMI or BMI change over time. Second, the longitudinal change and baseline status of each MetS component also predicted the development of hepatic steatosis. In other words, the development of obesity, high TG, or high FG during follow-up, regardless of baseline BMI and other measurements, predicted the incidence of hepatic steatosis. Moreover, obesity, high TG, or high FG at baseline, irrespective of other MetS components, were risk factors for the development of hepatic steatosis. These findings are meaningful in terms of revealing the significance of baseline status as well as longitudinal change of MetS and its components as risk factors for the development of hepatic steatosis. These relationships have been little studied because of the paucity of longitudinal studies regarding these relationships. Insulin resistance has been traditionally considered as a plausible potential mechanism for the co-existence of MetS and NAFLD [1—3,19]. The accumulation of TG within
BMI ≥ 25 kg/m2
Discussion
Table 3
variables also showed very similar associations. For subjects who initially did not have any MetS components but eventually had at least one MetS component on the follow-up assessment, the risk for the development of hepatic steatosis was nearly 3.2-fold higher in men and 8.4-fold higher in women (Table 2). During follow-up, the percentage of people with continued obesity, high BP, high TG, low HDL-C, or high FG were 19.2%, 12.0%, 6.5%, 8.9%, and 8.2%, respectively. Compared with the reference group, hepatic steatosis was more likely to occur among the subjects who met the criteria of obesity, high TG, or high FG on follow-up assessment, irrespective of age, sex, and baselinemeasurements (BMI, TG, FG, systolic BP, and HDL-C) (Table 3).
K. Lee
Association between categories of each metabolic syndrome (MetS) component at baseline and follow-up and the development of hepatic steatosis.
e222
Metabolic syndrome predicts NAFLD the hepatocytes, which results from an imbalance in the uptake, synthesis, export, and oxidation of fatty acids derived from several different sources, such as dietary fat, adipocytes, and de novo hepatic lipogenesis, contributes to the development of hepatic steatosis. Insulin resistance may contribute to hepatic steatosis indirectly by excess free fatty acid (FFA) delivery to the liver via decreased inhibition of lipolysis and directly by increasing de novo lipogenesis. In the current longitudinal study, subjects who had at least one or two MetS components were at higher risk of hepatic steatosis. In this regard, ultrasonographically detected hepatic steatosis appears to be a delayed or progressed manifestation of insulin resistance, while the condition of having MetS component may be a relatively early manifestation. When compared with a similar cohort study, there is discrepancy in the magnitude of associations between presence of MetS at baseline and the development of hepatic steatosis. Hamaguchi et al. [12] reported that the presence of MetS at baseline predicted the development of NAFLD with odds ratios of 4.0 for men and 11.2 for women, after adjustment for age, drinking states, and weight gain. In contrast, in this study, the hazard ratios (95% confidence interval) were 2.9 (2.1—3.9) for men and 3.8 (2.0—7.2) for women, after adjustment for age and weight gain. These differences may be due to differences in the characteristics of participants, such as the prevalence and incidence of the two conditions (which may be determined by ethnic background, prevalence of obesity, lifestyle and diagnostic tools) and due to differences in statistical analyses methods. These differences in studies also suggest that more longitudinal studies in other populations are needed to generalize these findings. However, several limitations should be considered. First, studies have reported that the presence of metabolic syndrome carried a high risk of hepatic fibrosis [20]; however, the current study did not evaluate of the association between MetS and the severity of hepatic steatosis. Second, although individuals with normal range of ALT and GGT at baseline were included to exclude those excessively consumed alcohol, alcohol effects on the relationship may not be completely separated. Third, there could be possible misclassification of hepatic steatosis because of the wide range in sensitivity and specificity of ultrasonography in detecting hepatic steatosis between observers [21]. Fourth, this study used data generated from adults referring to a university hospital; so they have to be considered as belonging to a clinic, hospital based cohort, not even a population-based cohort. There could have
e223 been a selection bias, because of different characteristics among people referring to this health center and common population. Finally, other conditions such as management of metabolic syndrome and alcohol consumption that would influence the development of MetS and hepatic steatosis during the follow-up may not have been completely accounted for in the analyses. In conclusion, this study among Korean adults with normal level of GGT and ALT demonstrates that individuals who had one or two components of MetS at baseline as well as those who initially did not have any MetS components but developed at least one MetS component on follow-up, regardless of their BMI, could be classified as a high risk group for the future development of hepatic steatosis.
Conflict of interest statement None declared.
References [1] Utzschneider KM, Kahn SE. Review: the role of insulin resistance in nonalcoholic fatty liver disease. J Clin Endocrinol Metab 2006;91:4753—61. [2] Almeda-Valdes P, Cuevas-Ramos D, Aguilar-Salinas CA. Metabolic syndrome and non-alcoholic fatty liver disease. Ann Hepatol 2009;8(Suppl. 1):S18—24. [3] Rector RS, Thyfault JP, Wei Y, Ibdah JA. Non-alcoholic fatty liver disease and the metabolic syndrome: an update. World J Gastroenterol 2008;14:185—92. [4] Caballeria L, Auladell MA, Toran P, Miranda D, Aznar J, Pera G, et al. Prevalence and factors associated with the presence of non alcoholic fatty liver disease in an apparently healthy adult population in primary care units. BMC Gastroenterol 2007;7:41. [5] Watanabe S, Yaginuma R, Ikejima K, Miyazaki A. Liver diseases and metabolic syndrome. J Gastroenterol 2008;43:509—18. [6] Angulo P. GI epidemiology: nonalcoholic fatty liver disease. Aliment Pharmacol Ther 2007;25:883—9. [7] Adams LA, Lymp JF, St Sauver J, Sanderson SO, Lindor KD, Feldstein A, et al. The natural history of nonalcoholic fatty liver disease: a population-based cohort study. Gastroenterology 2005;129:113—21. [8] Targher G, Bertolini L, Poli F, Rodella S, Scala L, Tessari R, et al. Nonalcoholic fatty liver disease and risk of future cardiovascular events among type 2 diabetic patients. Diabetes 2005;54:3541—6. [9] Lee K, Sung JA, Kim JS, Park TJ. The roles of obesity and gender on the relationship between metabolic risk factors and non-alcoholic fatty liver disease in Koreans. Diabetes Metab Res Rev 2009;25:150—5. [10] Hsiao PJ, Kuo KK, Shin SJ, Yang YH, Lin WY, Yang JF, et al. Significant correlations between severe fatty liver and risk factors for metabolic syndrome. J Gastroenterol Hepatol 2007;22:2118—23.
e224
K. Lee
[11] Cortez-Pinto H, Camilo ME, Baptista A, De Oliveira AG, De Moura MC. Non-alcoholic fatty liver: another feature of the metabolic syndrome? Clin Nutr 1999;18:353—8. [12] Hamaguchi M, Kojima T, Takeda N, Nakagawa T, Taniguchi H, Fujii K, et al. The metabolic syndrome as a predictor of nonalcoholic fatty liver disease. Ann Intern Med 2005;143:722—8. [13] Persson J, Magnusson PH. Comparison between different methods of detecting patients with excessive consumption of alcohol. Acta Med Scand 1988;223:101—9. [14] Moussavian SN, Becker RC, Piepmeyer JL, Mezey E, Bozian RC. Serum gamma-glutamyl transpeptidase and chronic alcoholism. Influence of alcohol ingestion and liver disease. Dig Dis Sci 1985;30:211—4. [15] Salaspuro M. Characteristics of laboratory markers in alcohol-related organ damage. Scand J Gastroenterol 1989;24:769—80. [16] Saverymuttu SH, Joseph AE, Maxwell JD. Ultrasound scanning in the detection of hepatic fibrosis and steatosis. Br Med J (Clin Res Ed) 1986;292:13—5.
[17] World Health Organization IAftSoO, Force IOT. The AsiaPacific perspective: redefining obesity and its treatment. Sydney Health Communications; 2000. [18] Alberti KG, Zimmet P, Shaw J. Metabolic syndrome— –a new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabet Med 2006;23:469—80. [19] Musso G, Gambino R, Bo S, Uberti B, Biroli G, Pagano G, et al. Should nonalcoholic fatty liver disease be included in the definition of metabolic syndrome? A cross-sectional comparison with Adult Treatment Panel III criteria in nonobese nondiabetic subjects. Diabetes Care 2008;31:562—8. [20] Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M, Manini R, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology 2003;37: 917—23. [21] Perez NE, Siddiqui FA, Mutchnick MG, Dhar R, Tobi M, Ullah N, et al. Ultrasound diagnosis of fatty liver in patients with chronic liver disease: a retrospective observational study. J Clin Gastroenterol 2007;41:624—9.
Available online at www.sciencedirect.com
ORIGINAL ARTICLE
Metabolic syndrome predicts the incidence of hepatic steatosis in Koreans Kayoung Lee ∗ Department of Family Medicine, Busan Paik Hospital, Inje University College of Medicine, 633-165 Kaegum-dong, Busan Jin-Gu, Busan 614-735, South Korea Received 5 November 2009 ; received in revised form 30 January 2010; accepted 2 February 2010
KEYWORDS Metabolic syndrome; Hepatic steatosis; Incidence
∗
Summary Problem: It has not been well elucidated whether the development of metabolic syndrome and its components predicts the incidence of hepatic steatosis. Methods: A cohort of 1705 apparently healthy Korean adults (954 men and 751 women, 43.6 ± 8.5 years old) without ultrasonographically defined hepatic steatosis and with normal serum gamma-glutamyl-transpeptidase and alanine aminotransferase was followed from 2004 to 2007. Metabolic syndrome was defined as the presence of at least three of the following components: obesity (BMI ≥ 25.0 kg/m2 ), high blood pressure, elevated levels of triglycerides and fasting glucose, and low level of high-density lipoprotein cholesterol. The outcome was ultrasonographically diagnosed hepatic steatosis. The analyses were conducted using the Cox proportional hazards model and time-dependent Cox model. Results: A total of 226 individuals developed hepatic steatosis during 3716 person—years of follow-up. The presence of one to two metabolic syndrome components at baseline significantly predicted the development of hepatic steatosis. Metabolic syndrome itself, having ≥1 metabolic syndrome components, and maintenance of metabolic syndrome during follow-up significantly increased the risk (hazard ratio 2.0—4.1 for men, 3.4—10.8 for women) after adjustment for the followup period, age and BMI at baseline or updated during follow-up. Occurrence of obesity, high triglycerides or high fasting glucose during follow-up significantly predicted the development of hepatic steatosis, even after adjustment for metabolic syndrome components at baseline. Conclusions: The presence at baseline and the development of metabolic syndrome during follow-up were risk factors for ultrasonographically detected hepatic steatosis. © 2010 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
Tel.: +82 51 890 6229; fax: +82 51 894 7554. E-mail addresses: [email protected], [email protected].
1871-403X/$ — see front matter © 2010 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2010.02.004
e218
Introduction Hepatic steatosis is a common hepatic disorder characterized by fat accumulation in the liver seen in individuals who excessively drink alcohol (alcohol-related fatty liver) or in those without history of relevant alcohol consumption (nonalcoholic fatty liver disease, NAFLD) [1,2]. NAFLD is frequently associated with conditions manifested by insulin resistance, such as obesity, hypertension, dyslipidemia, type 2 diabetes, and metabolic syndrome (MetS) [3—5]. There is evidence that NAFLD is related to higher mortality caused by liver-related complications [6,7] and to an elevated risk of cardiovascular disease [8]. While several cross-sectional studies have demonstrated the co-existence of metabolic syndrome and NAFLD [9—11], only one cohort study addressed the association between the presence of MetS at baseline and the future development of NAFLD [12]. However, the use of self-reported information for alcohol consumed may be prone to error and may bias the differentiation between NAFLD and alcohol-related liver disease. Studies have reported that self-reports appear to underestimate alcohol consumption in a sizable proportion of patients in clinical studies [13]. In contrast, serum gamma-glutamyl-transpeptidase (GGT) and alanine aminotransferase (ALT) has been shown to be sensitive markers of liver disease related with alcohol consumption [14,15]. Therefore, the clarification of relationship between MetS and incidence of hepatic steatosis in a prospective study among participants with normal range of ALT and GGT is helpful to extend knowledge about the temporal relationships between the two conditions among individuals after excluding alcohol-related liver damage. This study was carried out to evaluate the role of MetS and its components, at baseline and during follow-up, as risk factors for the development of ultrasonographically diagnosed hepatic steatosis among Korean men and women with normal range of ALT and GGT.
Methods Study subjects This study used data generated from 7553 Korean adults (4336 men and 3217 women, aged 20 years or older) who visited the Center of Health Promotion at Inje University, Busan Paik Hospital in 2004 to have health examinations. Subjects with evidence of excessive alcohol intake (≥20 g/day), abnormal
K. Lee level of GGT (>50 IU/L) and ALT (> 45 IU/L), positive seromarkers for hepatitis B or C, biliary disease, liver cirrhosis, or malignant disease were excluded through laboratory tests and ultrasonography at baseline. Alcohol intake was assessed using two open questions; ‘‘How often did you have a drink containing alcohol per week in the past 6 months?’’ and ‘‘How many glasses did you have on a typical day when you were drinking in the past 6 months? In the second question, one glass of alcoholic beverage was assumed to contain 10 g of alcohol. From these two questions, we calculated an average daily intake of alcohol. A total of 5986 individuals were eligible for the study at baseline, and of these, 1705 apparently healthy individuals (954 men and 751 women, 43.6 ± 8.5 years old) voluntarily received at least two evaluations and without specific regular interval but at least after one year from baseline examination, including abdominal ultrasonography, biochemical tests for liver and metabolic function, and measurements of body mass index (BMI), from 2004 until December of 2007. Compared to those excluded in the analyses (N = 4281), the individuals finally included in analyses were younger, and less likely to be women and fulfilled with the criteria of MetS components.
Measurements Health examinations performed at baseline and during follow-up were abdominal ultrasonography, biochemical tests for liver and metabolic function, blood pressure (BP) measurement, and anthropometric measurements. Hepatic steatosis was diagnosed using abdominal B-mode ultrasonography (LOGIQ 7 with a 4 MHz transducer; GE Healthcare, Wauwatosa, WI, USA). Ultrasonography was performed by experienced radiologists who did not have access to subjects’ clinical and laboratory test findings. Hepatic steatosis was originally classified from mild to severe levels but defined in current analyses when at least the criteria of mild hepatic steatosis were satisfied: slight increase in hepatic echogenicity and differences between hepatic and renal echogenicity, and relative preservation of echoes from the walls of the portal vein [16]. Body weight and height were measured in subjects clothed in a light gown without shoes; these measurements were used to calculate the BMI (kg/m2 ). A BMI of 25 kg/m2 or greater, which has been proposed for Asians as indicative of obesity, was used to define the presence of obesity [17]. Blood pressure measurements were performed with a standard manual sphygmomanometer while participants were in a sitting position. Antecubital venous blood samples were
Metabolic syndrome predicts NAFLD taken from all subjects after a 12-h overnight fast. The levels of high-density lipoprotein cholesterol (HDL-C), triglycerides (TGs), fasting glucose (FG), alanine aminotransferase (ALT), and gammaglutamyl-transpeptidase (GGT) were measured using a biochemistry autoanalyzer (Toshiba 200FR; Toshiba, Tokyo, Japan). The following criteria were used to define abnormalities of the MetS components: BP ≥ 130/85 mmHg; fasting glucose (FG) ≥ 100 mg/dL; high-density lipoprotein cholesterol (HDL-C) < 40 mg/dL for men or < 50 mg/dL for women; and TG ≥ 150 mg/dL [18]. Individuals with at least three abnormal MetS components, among the four components mentioned above and overall obesity (abdominal obesity was not used, because waist circumference was not assessed). The ethics committee of Inje University, Busan Paik Hospital approved this study.
Statistical analyses The 2 -test and t-test were used to assess the association between the baseline characteristics and the development of hepatic steatosis. Multivariate Cox proportional hazards modeling of the development of hepatic steatosis was performed to investigate relationships with each MetS component at baseline, after adjustment for follow-up duration, other MetS components, sex, and age at baseline. The combinations of the presence of abnormal MetS components and MetS at baseline and at follow-up assessments for each subject were categorized into four groups: no MetS at both assessments, change from non-MetS to MetS, change from MetS to no MetS, and having MetS at both tests. The follow-up assessment was defined as the assessment performed at a follow-up visit when hepatic steatosis was diagnosed for incident cases or the final assessment for non-cases. The Cox proportional hazards model and time-dependent Cox regression model were used separately in men and women to assess the risk for developing hepatic steatosis across the four categories of MetS, after adjustment for age, and BMI at baseline. In the time-dependent Cox regression analysis, the MetS categories and the change in covariates over follow-up time were modeled as time-dependent variables. The follow-up duration was calculated as the interval (as years) from the first visit until the time of hepatic steatosis development or until the final assessment, for each individual. Cox proportional hazard model was also used to assess the risk for incident hepatic steatosis with each MetS component category, after adjustment for age, sex, and various baseline-measurements (BMI, systolic BP, TG, HDL-C, and FG). Statistical significance
e219 was indicated when P < 0.05. All statistical analyses were performed using SPSS 17.0 (Release 17.0.0 (23rd, Apr. 2008); SPSS Inc., Chicago, IL, USA).
Results Longitudinal change of metabolic syndrome and hepatic steatosis Of the 1705 subjects without hepatic steatosis at baseline, the prevalence of MetS changed from 7.4% at baseline to 7.2% on follow-up: obesity, from 22.6% to 24.3%; high BP, from 19.8% to 18.0%; high TG, from 11.7% to 14.6%; low HDL-C, from 21.9% to 11.6%; and high FG, from 13.0% to 14.8%. During the follow-up, 88.8% of subjects remained free of MetS, while 3.4% of subjects (4.7% of men, 1.7% of women) continued to have MetS. For the 4.0% of subjects (4.6% of men, 3.3% of women) who had MetS at study entry, the MetS criteria were not more fulfilled during follow-up. In contrast, 3.8% of subjects (4.8% of men, 2.4% of women) without MetS at baseline developed incident MetS during the followup time. A total of 226 individuals (167 men and 59 women) developed hepatic steatosis during 3716 (2074 for men and 1642 for women) person—years of follow-up.
Risk for the development of hepatic steatosis by metabolic syndrome component at baseline Compared with the reference group, subjects with high BMI, high TG, or high FG at baseline were more likely to develop hepatic steatosis, after adjusting for the presence of other MetS components, age, and sex. The risk of developing hepatic steatosis was almost 6-fold higher in subjects who had MetS at baseline and was 2.4-fold higher even for those with one or two components at baseline (Table 1).
Risk for the development of hepatic steatosis by metabolic syndrome category during the follow-up time Table 2 shows that the risk for developing hepatic steatosis significantly increased with the development of MetS. Compared with those without MetS during follow-up, the presence of MetS at the follow-up assessment predicted the development of hepatic steatosis, after adjustment for age and BMI at baseline. The Cox regression models including the four categories of MetS, changes in BMI and other covariates over time as time-dependent
e220
Table 1
The comparison of baseline characteristics by category of metabolic syndrome. Overall, N = 1705
FL (−), N = 1479
FL (+), N = 226
Mean ± standard deviation, N (%) Age and sex-adjusted
Multivariate† — —
43.6 ± 8.5 954 (56.0)
43.3 ± 8.5 787 (53.2)
46.2 ± 8.3 167 (73.9)
— —
385 (22.6) 338 (19.8) 199 (11.7) 373 (21.9) 222 (13.0)
283 (19.1) 280 (18.9) 137 (9.3) 302 (20.4) 168 (11.4)
102 (45.1) 58 (25.7) 62 (27.4) 71 (31.4) 54 (23.9)
2.74 1.14 2.67 1.76 1.74
Number of MetS components at baseline*2 0 743 (43.6) 1—2 835 (49.0) 3—5 127 (7.4)
697 (47.1) 705 (47.7) 77 (5.2)
46 (20.4) 130 (57.5) 50 (22.1)
1.0 2.37 (1.69—3.33) 5.91 (3.93—8.89)
Age (y) Male*1
*1
MetS component at baseline BMI ≥ 25 kg/m2*2 BP ≥ 130/85 mmHg*2 TG ≥ 150 mg/dL*2 HDL-C < 40 (M) 50 (F) mg/dL*2 FG ≥ 100 mg/dL*2
HR (95% CI)
(2.11—3.57) (0.84—1.56) (1.99—3.58) (1.32—2.33) (1.26—2.40)
2.46 0.99 2.10 1.23 1.61
(1.88—3.22) (0.72—1.34) (1.52—2.89) (0.91—1.68) (1.16—2.22)
— —
Abbreviations: MetS, metabolic syndrome; BMI, body mass index; BP, blood pressure; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; FG, fasting glucose. * P < 0.05 using 1 t-test or 2 chi-square test. † Using Cox proportional hazard models after adjustment for age, sex, and other MetS components at baseline.
K. Lee
Association of metabolic syndrome (MetS) categories at baseline and follow-up with the development of hepatic steatosis. Person—years
Incident case (N, %)
Age-adjusted HR (95% CI)
Multivariate HR (95% CI)*
HR adjusted with time-dependent variables (95% CI)†
Male (N = 954) Both non-MetS (N = 819) From MetS to non-MetS (N = 44) From non—MetS to MetS (N = 46) Both MetS (N = 45)
1801 98 92 83
115 (14.0) 15 (34.1) 16 (34.8) 21 (46.7)
1.00 2.38 (1.39—4.08) 2.76 (1.64—4.66) 4.17 (2.61—6.64)
1.00 1.60 (0.91—2.81) 1.96 (1.13—3.38) 2.36 (1.40—4.01)
1.00 2.82 (1.64—4.85) 1.84 (1.01—3.35) 4.11 (2.58—6.54)
Female (N = 751) Both non-MetS (N = 695) From MetS to non-MetS (N = 25) From non-MetS to MetS (N = 18) Both MetS (N = 13)
1527 53 35 27
39 (5.6) 6 (24.0) 6 (33.3) 8 (61.5)
1.00 3.40 (1.41—8.22) 5.73 (2.37—13.86) 7.05 (2.98—16.71)
1.00 2.22 (0.90—5.43) 4.26 (1.77—10.26) 3.42 (1.39—8.36)
1.00 3.48 (1.43—8.43) 5.00 (2.06—12.16) 5.87 (2.35—14.69)
1.0 3.97 (2.52—6.24)
1.0 3.19 (1.98—5.13)
1.0 3.63 (2.31—5.72)
1.0 11.28 (4.40—28.89)
1.0 8.40 (3.16—22.33)
1.0 10.82 (4.24—27.62)
Male without any one MetS component at baseline (N = 865) No MetS component on f/u (N = 373) 855 MetS component≥ 1 on f/u (N = 492) 1038 Female without any one MetS component at baseline (N = 713) No MetS component on f/u (N = 438) 958 MetS component ≥ 1 on f/u (N = 275) 604
23(6.2) 108(22.0) 5 (1.1) 40 (14.5)
Metabolic syndrome predicts NAFLD
Table 2
*
Using Cox proportional hazard models after adjustment for age, follow-up duration, and BMI at baseline. Using extended Cox proportional hazard models with MetS category as a time-dependent variable and adjustment for age, follow-up duration, and change of BMI during the follow-up time as time-dependent variables. †
e221
Units are hazard ratio (95% confidence interval) for developing hepatic steatosis with each MetS component category using Cox proportional hazard models after controlling for age, sex, and baseline-measurement values (BMI, systolic blood pressure, fasting glucose, triacylglycerides, and high-density lipoprotein cholesterol).
1.0 1.41 (0.79—2.51) 1.78 (1.18—2.70) 1.90 (1.18—3.07) 80.4 4.8 6.6 8.2 1.0 0.52 (0.33—0.81) 1.75 (0.88—3.48) 0.66 (0.41—1.06) 75.9 13.1 2.1 8.9 1.0 1.53 (0.91—2.55) 2.08 (1.41—3.05) 2.20 (1.35—3.60) 80.2 5.2 8.1 6.5 1.0 0.73 (0.44—1.20) 1.11 (0.72—1.73) 0.96 (0.59—1.56) 74.2 7.8 6.0 12.0 1.0 0.78 (0.33—1.87) 2.77 (1.71—4.47) 1.92 (1.15—3.19)
HR (95% CI) % HR (95% CI) % HR (95% CI) % HR (95% CI)
HR (95% CI) % %
72.3 3.3 5.1 19.2 Category Both non-MetS From MetS to non-MetS From non-MetS to MetS Both MetS
FG ≥ 100 mg/dL HDL-C < 40 (M) 50 (F) mg/dL TG ≥ 150 mg/dL BP ≥ 130/85 mmHg
This longitudinal study among individuals with normal level of ALT and GGT at baseline confers some important findings on the relationship between MetS and hepatic steatosis. First, the longitudinal change in MetS itself was a significant risk factor of the incidence of hepatic steatosis. Compared with the reference group, the incidence of hepatic steatosis was 1.8—4.1-fold higher and 3.4—5.9-fold higher for men and women, respectively, who had MetS on the follow-up assessment, regardless of baseline MetS status, baseline BMI or BMI change over time. Even those who initially did not have any MetS components but had one MetS component on follow-up were at higher risk for future development of hepatic steatosis, regardless of baseline BMI or BMI change over time. Second, the longitudinal change and baseline status of each MetS component also predicted the development of hepatic steatosis. In other words, the development of obesity, high TG, or high FG during follow-up, regardless of baseline BMI and other measurements, predicted the incidence of hepatic steatosis. Moreover, obesity, high TG, or high FG at baseline, irrespective of other MetS components, were risk factors for the development of hepatic steatosis. These findings are meaningful in terms of revealing the significance of baseline status as well as longitudinal change of MetS and its components as risk factors for the development of hepatic steatosis. These relationships have been little studied because of the paucity of longitudinal studies regarding these relationships. Insulin resistance has been traditionally considered as a plausible potential mechanism for the co-existence of MetS and NAFLD [1—3,19]. The accumulation of TG within
BMI ≥ 25 kg/m2
Discussion
Table 3
variables also showed very similar associations. For subjects who initially did not have any MetS components but eventually had at least one MetS component on the follow-up assessment, the risk for the development of hepatic steatosis was nearly 3.2-fold higher in men and 8.4-fold higher in women (Table 2). During follow-up, the percentage of people with continued obesity, high BP, high TG, low HDL-C, or high FG were 19.2%, 12.0%, 6.5%, 8.9%, and 8.2%, respectively. Compared with the reference group, hepatic steatosis was more likely to occur among the subjects who met the criteria of obesity, high TG, or high FG on follow-up assessment, irrespective of age, sex, and baselinemeasurements (BMI, TG, FG, systolic BP, and HDL-C) (Table 3).
K. Lee
Association between categories of each metabolic syndrome (MetS) component at baseline and follow-up and the development of hepatic steatosis.
e222
Metabolic syndrome predicts NAFLD the hepatocytes, which results from an imbalance in the uptake, synthesis, export, and oxidation of fatty acids derived from several different sources, such as dietary fat, adipocytes, and de novo hepatic lipogenesis, contributes to the development of hepatic steatosis. Insulin resistance may contribute to hepatic steatosis indirectly by excess free fatty acid (FFA) delivery to the liver via decreased inhibition of lipolysis and directly by increasing de novo lipogenesis. In the current longitudinal study, subjects who had at least one or two MetS components were at higher risk of hepatic steatosis. In this regard, ultrasonographically detected hepatic steatosis appears to be a delayed or progressed manifestation of insulin resistance, while the condition of having MetS component may be a relatively early manifestation. When compared with a similar cohort study, there is discrepancy in the magnitude of associations between presence of MetS at baseline and the development of hepatic steatosis. Hamaguchi et al. [12] reported that the presence of MetS at baseline predicted the development of NAFLD with odds ratios of 4.0 for men and 11.2 for women, after adjustment for age, drinking states, and weight gain. In contrast, in this study, the hazard ratios (95% confidence interval) were 2.9 (2.1—3.9) for men and 3.8 (2.0—7.2) for women, after adjustment for age and weight gain. These differences may be due to differences in the characteristics of participants, such as the prevalence and incidence of the two conditions (which may be determined by ethnic background, prevalence of obesity, lifestyle and diagnostic tools) and due to differences in statistical analyses methods. These differences in studies also suggest that more longitudinal studies in other populations are needed to generalize these findings. However, several limitations should be considered. First, studies have reported that the presence of metabolic syndrome carried a high risk of hepatic fibrosis [20]; however, the current study did not evaluate of the association between MetS and the severity of hepatic steatosis. Second, although individuals with normal range of ALT and GGT at baseline were included to exclude those excessively consumed alcohol, alcohol effects on the relationship may not be completely separated. Third, there could be possible misclassification of hepatic steatosis because of the wide range in sensitivity and specificity of ultrasonography in detecting hepatic steatosis between observers [21]. Fourth, this study used data generated from adults referring to a university hospital; so they have to be considered as belonging to a clinic, hospital based cohort, not even a population-based cohort. There could have
e223 been a selection bias, because of different characteristics among people referring to this health center and common population. Finally, other conditions such as management of metabolic syndrome and alcohol consumption that would influence the development of MetS and hepatic steatosis during the follow-up may not have been completely accounted for in the analyses. In conclusion, this study among Korean adults with normal level of GGT and ALT demonstrates that individuals who had one or two components of MetS at baseline as well as those who initially did not have any MetS components but developed at least one MetS component on follow-up, regardless of their BMI, could be classified as a high risk group for the future development of hepatic steatosis.
Conflict of interest statement None declared.
References [1] Utzschneider KM, Kahn SE. Review: the role of insulin resistance in nonalcoholic fatty liver disease. J Clin Endocrinol Metab 2006;91:4753—61. [2] Almeda-Valdes P, Cuevas-Ramos D, Aguilar-Salinas CA. Metabolic syndrome and non-alcoholic fatty liver disease. Ann Hepatol 2009;8(Suppl. 1):S18—24. [3] Rector RS, Thyfault JP, Wei Y, Ibdah JA. Non-alcoholic fatty liver disease and the metabolic syndrome: an update. World J Gastroenterol 2008;14:185—92. [4] Caballeria L, Auladell MA, Toran P, Miranda D, Aznar J, Pera G, et al. Prevalence and factors associated with the presence of non alcoholic fatty liver disease in an apparently healthy adult population in primary care units. BMC Gastroenterol 2007;7:41. [5] Watanabe S, Yaginuma R, Ikejima K, Miyazaki A. Liver diseases and metabolic syndrome. J Gastroenterol 2008;43:509—18. [6] Angulo P. GI epidemiology: nonalcoholic fatty liver disease. Aliment Pharmacol Ther 2007;25:883—9. [7] Adams LA, Lymp JF, St Sauver J, Sanderson SO, Lindor KD, Feldstein A, et al. The natural history of nonalcoholic fatty liver disease: a population-based cohort study. Gastroenterology 2005;129:113—21. [8] Targher G, Bertolini L, Poli F, Rodella S, Scala L, Tessari R, et al. Nonalcoholic fatty liver disease and risk of future cardiovascular events among type 2 diabetic patients. Diabetes 2005;54:3541—6. [9] Lee K, Sung JA, Kim JS, Park TJ. The roles of obesity and gender on the relationship between metabolic risk factors and non-alcoholic fatty liver disease in Koreans. Diabetes Metab Res Rev 2009;25:150—5. [10] Hsiao PJ, Kuo KK, Shin SJ, Yang YH, Lin WY, Yang JF, et al. Significant correlations between severe fatty liver and risk factors for metabolic syndrome. J Gastroenterol Hepatol 2007;22:2118—23.
e224
K. Lee
[11] Cortez-Pinto H, Camilo ME, Baptista A, De Oliveira AG, De Moura MC. Non-alcoholic fatty liver: another feature of the metabolic syndrome? Clin Nutr 1999;18:353—8. [12] Hamaguchi M, Kojima T, Takeda N, Nakagawa T, Taniguchi H, Fujii K, et al. The metabolic syndrome as a predictor of nonalcoholic fatty liver disease. Ann Intern Med 2005;143:722—8. [13] Persson J, Magnusson PH. Comparison between different methods of detecting patients with excessive consumption of alcohol. Acta Med Scand 1988;223:101—9. [14] Moussavian SN, Becker RC, Piepmeyer JL, Mezey E, Bozian RC. Serum gamma-glutamyl transpeptidase and chronic alcoholism. Influence of alcohol ingestion and liver disease. Dig Dis Sci 1985;30:211—4. [15] Salaspuro M. Characteristics of laboratory markers in alcohol-related organ damage. Scand J Gastroenterol 1989;24:769—80. [16] Saverymuttu SH, Joseph AE, Maxwell JD. Ultrasound scanning in the detection of hepatic fibrosis and steatosis. Br Med J (Clin Res Ed) 1986;292:13—5.
[17] World Health Organization IAftSoO, Force IOT. The AsiaPacific perspective: redefining obesity and its treatment. Sydney Health Communications; 2000. [18] Alberti KG, Zimmet P, Shaw J. Metabolic syndrome— –a new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabet Med 2006;23:469—80. [19] Musso G, Gambino R, Bo S, Uberti B, Biroli G, Pagano G, et al. Should nonalcoholic fatty liver disease be included in the definition of metabolic syndrome? A cross-sectional comparison with Adult Treatment Panel III criteria in nonobese nondiabetic subjects. Diabetes Care 2008;31:562—8. [20] Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M, Manini R, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology 2003;37: 917—23. [21] Perez NE, Siddiqui FA, Mutchnick MG, Dhar R, Tobi M, Ullah N, et al. Ultrasound diagnosis of fatty liver in patients with chronic liver disease: a retrospective observational study. J Clin Gastroenterol 2007;41:624—9.
Available online at www.sciencedirect.com
