Gender Difference in Hepatic Steatosis Index and Lipid Accumulation Product Ability to Predict Incident Metabolic Syndrome in the Historical Cohort of the Brisighella Heart Study

Abstract

Background:

Nonalcoholic fatty liver disease (NAFLD) and metabolic syndrome are strongly related from both an epidemiological and a pathological point of view. The main aim of our study was to evaluate if two validated indexes of NAFLD are able to predict the 4-year metabolic syndrome incidence in a large population sample of pharmacologically untreated subjects without metabolic syndrome, type 2 diabetes, history of alcohol abuse, or known liver diseases at the baseline.

Methods:

From the database of the Brisighella Heart Study, we selected a subsample of 824 pharmacologically untreated subjects (male, 401; female, 423) without metabolic syndrome, type 2 diabetes, alcohol abuse, or known liver diseases at the 2004 survey and revisited in 2008. The Hepatic Steatosis Index (HSI) and Lipid Accumulation Product (LAP) values were calculated for everyone to evaluate their predictive role for metabolic syndrome through a Cox-regression analysis adjusted by metabolic syndrome components.

Results:

We observed 46 new cases of metabolic syndrome (male, 25; female, 21) with a cumulative incidence of 5.6% (1.4% per year): 6.2% in men (1.5% per year), 4.9% in women (1.2% per year). In women, the 4-year metabolic syndrome predictors (R ²=0.680) were age [odds ratio (OR) 1.13, 95% confidence interval (CI) 1.12–1.15], HSI (OR 1.20, 95% CI 1.15–1.26), and lnLAP (OR 1.16, 95% CI 1.14–1.17), whereas in men they were (R ²=0.554) age (OR 1.13, 95% CI 1.11–1.14) and lnLAP (OR 1.17, 95% CI 1.15–1.18).

Conclusion:

HSI and LAP differently predict the 4-year incidence of metabolic syndrome. In women, both indexes can be considered significant predictors, whereas in men only LAP can be considered predictive.

Introduction

N onalcoholic fatty liver disease (NAFLD) is a major health problem, involving about 5%–40% of Asian populations¹ and 20%–30% of Western populations, particularly amongst obese subjects.^2,3 It is known that gender influences the prevalence of NAFLD, but data are still controversial, with earlier studies supporting a higher predisposition in women,⁴ whereas more recent ones suggest approximately a similar prevalence between sexes.^5,6 NAFLD is defined as the deposition of triglycerides (TGs) in liver >5% of total liver weight, with an alcohol consumption <10 grams daily.⁷ It includes a wide spectrum of chronic liver diseases representing a risk factor for increased all-cause mortality, liver-related death, type 2 diabetes, and cardiovascular diseases,^8,9 either independently, suggesting a specific role of hepatic fat in atherosclerosis,^10,11 or according to the high prevalence of associated risk factors (in particular obesity and insulin resistance).

Insulin-resistance seems to be the common pathophysiological feature among NAFLD, type 2 diabetes, and cardiovascular diseases, promoting central fat accumulation and inducing a chronic low-grade inflammatory status. The consequent alterations observed in glucose, fatty acid and lipoprotein metabolism, and increased oxidative stress, together with a modified adipokine profile, endothelial dysfunction, and accelerated atherosclerotic process,¹² make NAFLD the hepatic expression of the metabolic syndrome.¹³ Moreover, epidemiological data support the hypothesis that metabolic syndrome is an independent risk factor for NAFLD, although it is difficult to distinguish which one is the cause and which the effect.¹⁴ The main risk factors for NAFLD are also components of metabolic syndrome, and several observations suggest that NAFLD may develop independently from metabolic syndrome¹⁵ or it may also predict metabolic syndrome.^16,17 However, metabolic syndrome components have a different distribution in men and women, with a significantly higher prevalence of central obesity and type 2 diabetes in women¹⁸ and conversely greater visceral fat with a more severe metabolic phenotype in men,¹⁹ and also a supposed but still controversial different gender prevalence of NAFLD as well.²⁰

In this context, the main aim of our study was to evaluate if two validated indexes of NAFLD are able to predict the 4-year metabolic syndrome incidence in a large population sample of pharmacologically untreated subjects without metabolic syndrome, type 2 diabetes, history of alcohol abuse, or known liver diseases at the baseline. The secondary aim was to evaluate if the ability of NAFLD indexes to predict incident metabolic syndrome is different in men and women.

Methods

Brisighella Heart Study

The Brisighella Heart Study (BHS) is a prospective, population-based, longitudinal epidemiological investigation that involves 2939 randomly selected subjects (1491 men and 1448 women), aged 14–84 years, free of cardiovascular diseases at enrollment, resident in the northern Italian rural town of Brisighella. The study started in 1972²¹ and is still ongoing. Brisighella was originally selected as the site for the study because of the homogeneity of lifestyle among its residents, with a very low rate of migration. Subjects were clinically evaluated at baseline and every 4 years thereafter, collecting an extensive amount of clinical and laboratory data. Between 1986 and 1994, the study was included among the working party of the WHO European Risk Factors Co-ordinated Analysis (ERICA),²² the Risk Factors and Life Expectancy Project,²³ and the World Health Organization Countrywide Integrated Noncommunicable Disease Intervention (WHO-CINDI) project. Data derived from the survey of the BHS population were also included in the development of a new algorithm to calculate the risk of cardiovascular diseases in the Italian population (RISCARD 2000²⁴ and RISKARD 2005²⁵). More recently (2007), they were entered into the Advanced Diagnostic Support in Lipidology (ADSL) project.²⁶ The BHS protocol and its substudies have been approved by the Ethical Board of the University of Bologna, and all volunteers gave their signed consent to participate in the study. A detailed description of the protocol of the BHS and the full protocol have been largely described elsewhere.²⁷

All-cause mortality and morbidity, as well as the incidence of the main cardiovascular risk factors, were recorded throughout the entire study. The study design includes a database update, with regard to fatal and nonfatal new events [encoded following the International Classification of Diseases, 9^th edition (ICD-9)] every 3 months, and a complete medical checkup of participants every 4 years. The standard visit included an update of familial and personal history (with specific attention to lifestyle habits and pharmacological treatments), a physical examination (with specific attention to anthropometric measurements, blood pressure measuring, heart and breath rate evaluation), and the collection of fasting blood sample and electrocardiography. Smoking habit was investigated routinely through a questionnaire collecting data on previous and current cigarette consumption (number, years of smoking, years from cessation). Physical activity was evaluated by the Minnesota questionnaire.²⁰ Anthropometric measures were also available, particularly body mass index (BMI) and waist circumference (WC). Ematochemistry analyses were evaluated according to standardized methods by trained personnel and included total cholesterol (TC), low-density lipoprotein-cholesterol (LDL-C), total TGs, high-density lipoprotein-cholesterol (HDL-C), transaminases, creatinine, and creatinine phosphokinase.

NAFLD study

For the purpose of the present study, we considered a subsample of 824 pharmacologically untreated participants (male, 401; female, 423) free from metabolic syndrome, type 2 diabetes, alcohol abuse, or known cardiovascular and liver diseases at the 2004 population survey, who were re-examined at the 2008 survey. Liver diseases at the baseline were excluded on the basis of anamnestic data cross-matched with the electronic clinical forms of the subject general practitioners (GPs). Transaminases were considered normal at the baseline if less than 40 U/L. If higher, they were rechecked and subjects were investigated for viral hepatitis. Subjects positive for viral hepatitis were excluded from this study. For each participant we calculated the hepatic steatosis index (HSI) {calculated as 8×[alanine aminotransferase (ALT)/aspartate transaminase (AST) ratio]+BMI (+2 if woman;+2 if T2D)}²⁸ and the lipid accumulation product (LAP) {calculated as [WC (cm)−65]×TG (mmol/L) for men and [WC (cm)−58]×TG (mmol/L) for women}.²⁹

Statistical analysis

A full descriptive analysis of the investigated population sample at the 2004 and 2008 surveys was carried out according to gender. The predictive role of LSI and lnLAP for metabolic syndrome was tested by a Cox regression analysis adjusted for metabolic syndrome components (WC, SBP, DBP, FPG, TGs, HDL-C). A P level of 0.05 was considered significant for all tests. Statistical analyses were carried out with the help of the SPSS 19.0, Version for Windows.

Results

The main anagraphical, anthropometric, and hemodynamic characteristics of the selected population at 2004 and 2008 BHS surveys according to gender are summarized in Table 1. The pertinent laboratory parameters are reported in Table 2. At baseline, most of the investigated parameters were matched between genders. Only WC and LAP were significantly higher in men (P<0.05), whereas HDL-C and apolipoprotein AI (ApoAI) were higher in women (P<0.05). Overall, from 2004 to 2008, a nonsignificant increasing trend was registered in the levels of all the metabolic syndrome components. Waist circumference, diastolic blood pressure (DBP), serum uric acid (SUA), creatinine, and ALT increased significantly from 2004 to 2008 in men only (P<0.05), while fasting plasma glucose (FPG) significantly increased in both genders (P<0.05). We observed 46 new cases of metabolic syndrome (male, 25; female, 21) with a cumulative incidence of about 5.6% (1.4% per year): 6.2% in men (1.5% per year), 4.9% in women (1.2% per year). Considering the entire population, the best predictors of metabolic syndrome were age [odds ratio (OR) 1.13, 95% confidence interval (CI) 1.12–1.143, P<0.001), HSI (OR 1.17, 95% CI 1.14–1.21, P<0.001), and lnLAP (OR 1.16, 95% CI 1.15–1.17, P<0.001). In men, after data adjustment, the 4-year metabolic syndrome predictors (R ²=0.554) were age (OR 1.13, 95% CI 1.11–1.14), and lnLAP (OR 1.17, 95% CI 1.15–1.18). In women, the 4-year metabolic syndrome predictors (R ²=0.680) were age (OR 1.13, 95% CI 1.12–1.15), HSI (OR 1.20, 95% CI 1.15–1.26), and lnLAP (OR 1.16, 95% CI 1.14–1.17). HSI and LAP in men and women who developed or did not have metabolic syndrome are graphically reported in Fig. 1.

FIG. 1.

Baseline level of LSI (A) and LAP (B) in patients who developed or did not develop metabolic syndrome in men and women. LDL-C, low-density lipoprotein cholesterol; hsCRP, high-sensitivity C-reactive protein; MMP, matrix metalloproteinase.

Table 1.

Main Anagraphic, Anthropometric, and Hemodynamic Characteristics of the Selected Population During the 2004 and 2008 Brisighella Heart Study Surveys According to Gender

	Men (n=401)				Women (n=423)
	2004		2008		2004		2008
	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Age (years)	49.05	17.81	53.32^*	17.96	48.19	16.94	52.78^*	18.17
Body mass index (kg/m²)	25.31	4.84	26.82	3.78	25.23	4.62	26.08	4.11
Waist circumference (cm)	91.35	13.38	94.01^*	11.06	85.32^**	12.82	87.45^**	13.66
SBP (mmHg)	132.85	15.28	134.80	16.54	129.25	14.36	131.33	14.64
DBP (mmHg)	83.12	11.05	85.02^*	10.62	80.02	9.56	81.24	10.07
PP (mmHg)	49.47	4.51	50.61	4.63	49.18	5.12	50.39	4.97

P<0.05 vs. 2004; ^** P<0.05 vs. men.

SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure.

Table 2.

Main Anagraphic, Anthropometric, and Hemodynamic Characteristics of the Selected Population During the 2004 and 2008 Brisighella Heart Study Surveys According to Sex (N=824)

	Men				Women
	2004		2008		2004		2008
	Mean	SD	Mean	SD	Mean	SD	Mean	SD
FPG (mg/dL)	94.92	9.31	98.90^*	8.90	92.74	8.72	95.19^*	7.28
TC (mg/dL)	198.71	31.19	202.42	38.95	197.34	31.87	204.40	38.72
TG (mg/dL)	119.88	58.12	121.16	63.71	110.12	59.94	104.81	60.72
HDL-C (mg/dL)	43.89	9.21	43.14	9.87	50.44^**	8.61	49.12^**	10.79
LDL-C (mg/dL)	132.15	25.15	135.28	24.84	130.88	27.78	134.39	24.38
VLDL-C (mg/dL)	41.77	7.69	42.41	8.50	42.45	10.01	40.99	9.38
ApoB100 (mg/dL)	89.84	19.34	90.70	21.45	87.68	21.53	86.51	21.04
ApoAI (mg/dL)	132.78	23.46	133.53	25.66	157.18^**	24.22	152.02^**	27.79
SUA (mg/dL)	4.91	1.52	5.85^*	1.36	4.68	1.39	4.37	1.33
Creatinine (mg/dL)	0.92	0.16	1.19^*	0.18	0.89	0.23	0.85	0.19
AST (U/L)	24.34	5.54	26.04	8.29	21.78	7.77	22.29	8.23
ALT (U/L)	24.99	4.43	29.89^*	9.39	22.46	5.17	20.03^**	9.19
HSI	36.21	7.21	36.24	6.19	35.46	6.34	35.37	6.06
LAP	42.74	16.61	42.57	16.98	38.62^**	15.44	38.61	15.81

P<0.05 vs. 2004; ^** P<0.05 vs. men.

SD, standard deviation; FPG, fasting plasma glucose; TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; TGs, triglycerides; HDL-C, high-density lipoprotein cholesterol; Apo, apolipoprotein; SUA, serum uric acid; AST, aspartate aminotransferase; ALT, alanine aminotransferase; HSI, hepatic steatosis index; LAP, lipid accumulation product.

Discussion

HSI and LAP are already validated indexes of NAFLD. In this study, we demonstrated that LAP can also valuably predict the development of metabolic syndrome in subjects free from metabolic syndrome at baseline, in both genders. The discovery of these two indexes was the answer to the need to identify patients at high risk of NAFLD and to establish an appropriate screening program. Liver biopsy followed by an evaluation of the Kleiner histological NAFLD activity score is still the gold standard to ascertain the stage of the disease.³⁰ However, the invasive nature of the exam inevitably limits the diagnosis, suggesting exactly the discovery of noninvasive methods to diagnose NAFLD. Such methods include age, anthropometric measurements (BMI), and biochemical tests (glycemia, platelet count, albumin, and serum AST/ALT) in combination with ultrasound, computed tomography (CT) scan and magnetic resonance imaging (MRI).^31,32 However, for the impossibility of using ultrasonography (US) or CT as screening techniques of asymptomatic individuals due to their high cost, simple and noninvasive tests as HSI and LAP were developed to establish an appropriate NAFLD screening program, even if not validated on the basis of a liver biopsy. LAP,²² derived from the Dyonisos study,³ is an algorithm based on a combination of WC and fasting TGs. HSI, instead, is based on AST and ALT levels, BMI, presence or absence of diabetes, and sex.

Considering that NAFLD is the hepatic expression of metabolic syndrome, we investigated how HSI and LAP can be clinically used to predict incident metabolic syndrome in pharmacologically untreated men and women without metabolic syndrome, type 2 diabetes, history of alcohol abuse, or known cardiovascular and liver diseases at the baseline. It was observed that if lnLAP can be used as a significant predictor of incident metabolic syndrome in both sexes, this is not the same for HSI, which was a predictor only in women. The reason for these findings is not completely clear. The principal reason we found to explain why lnLAP is demonstrated to be a better predictor compared to HSI is that the lnLAP formula takes into consideration the measure of intra-abdominal fat depot (WC), whereas HSI considers BMI. WC, in fact, offers a better correlation with lipid risk variables, uric acid concentration, heart rate, and diabetes.^33,34 Moreover, compared to BMI, WC is a more reliable parameter of visceral adipose tissue, which with its great lipolytic activity is strictly related to the development of obesity and insulin resistance, well-known determinants of metabolic syndrome. On the other side, at this level of our research, we are not yet able to explain the reason why HSI is a significant predictor in women but not in men. Our study is the first one to use this index as a metabolic syndrome predictor in the Italian population, and the HSI formula has to be improved for this purpose.

We are aware of some relevant limitations of our study. The principal one is the lack of NAFLD evaluation using ultrasonography or more sophisticated techniques, although we have used validated NAFLD indexes. The reason for our choice is due to the healthy status of the population considered, which made estimating those techniques to be neither ethical nor cost-effective for screening purposes. Of course, other indexes could have been chosen, but the most part of them include serum γ-glutamyl transferase level, not available for all the investigated subjects in both considered Brisighella Heart Study surveys. Therefore, we have selected the participants according to the criteria listed in the Methods section and not in a randomized manner; this was necessary to select healthy subjects. For the same reason, the number of subjects considered is relatively few and it could have partly reduced the statistical power of the study, especially in men, who were mildly less prevalent in the studied population sample. However, the final sample was representative of the age-matched Brisighella general population distribution. Finally, other confounding variables might have influenced our results. The regression analyses were adjusted for a large number of parameters, so we are confident in our observation. To the best of our knowledge, this is the first study evaluating the ability of NAFLD-related indexes to predict the short-term incidence of metabolic syndrome in a large sample of men and women.

To conclude, on the basis of our observations, it seems that, although both HSI and lnLAP are validated indexes of NAFLD, only lnLAP can be clinically used as a significant short-term predictor of incident metabolic syndrome in both genders. HSI can be applicable in women but not in men. Until we have a longer follow-up and a validation of the indexes in study, including a liver biopsy, the relevance of these data remains relatively speculative.

Footnotes

Acknowledgment

This study was supported by institutional funding of the University of Bologna, Italy.

Author Disclosure Statement

No competing financial interests exist.

References

Amarapurkar

, Hashimoto

, Lesmana

et al. Asia-Pacific Working Party on NAFLD. How common is non-alcoholic fatty liver disease in the Asia–Pacific region and are there local differences? J Gastroenterol Hepatol, 2007; 22:788–793.

Browning

, Szczepaniak

, Dobbins

et al. Prevalence of hepatic steatosis in an urban population in the United States: Impact of ethnicity. Hepatology, 2004; 40:1387–1395.

Bedogni

, Miglioli

, Masutti

et al. Prevalence of and risk factors for nonalcoholic fatty liver disease: The Dionysos nutrition and liver study. Hepatology, 2005; 42:44–52.

Sanyal

. American Gastroenterological Association Clinical Practice Committee and American Association for the Study of Liver Diseases Practice Guidelines Committee. AGA technical review on nonalcoholic fatty liver disease. Gastroenterology, 2002; 123:1705–1725.

Chitturi

, Farrell

, Hashimoto

et al. Nonalcoholic fatty liver disease in the Asia-Pacific region: Definitions and overview of proposed guidelines. J Gastroenterol Hepatol, 2007; 22:778–787.

Farrell

, Larter

. Nonalcoholic fatty liver disease: From steatosis to cirrhosis. Hepatology, 2006; 43:S99–S112.

Farrell

. The liver and the waistline: Fifty years of growth. J Gastroenterol Hepatol, 2009; 24,Suppl 3:S105e18.

Adams

, Lymp

, St Sauver

et al. The natural history of non-alcoholic fatty liver disease: A population-based cohort study. Gastroenterology, 2005; 129:113–121.

Arslan

, Turkoglu

, Balcioglu

et al. Association between nonalcoholic fatty liver disease and coronary artery disease. Coron Artery Dis, 2007; 18:433–436.

10.

Yki-Järvinen

. Liver fat in the pathogenesis of insulin resistance and type 2 diabetes. Dig Dis, 2010; 28:203–209.

11.

Bhatia

, Curzen

, Calder

et al.

Non-alcoholic fatty liver disease: A new and important cardiovascular risk factor?

Eur Heart J, 2012; 33:1190–1200.

12.

Tarantino

, Caputi

. JNKs, insulin resistance and inflammation: A possible link between NAFLD and coronary artery disease. World J Gastroenterol, 2011; 17:3785–3794.

13.

Seo

, Cho

, Lee

et al.

Which metabolic syndrome criteria best predict the presence of non-alcoholic fatty liver disease?

Diabetes Res Clin Pract, 2012; 95:19–24.

14.

Hamaguchi

, Kojima

, Takeda

et al. The metabolic syndrome as a predictor of nonalcoholic fatty liver disease. Ann Intern Med, 2005; 143:722–728.

15.

Yilmaz

, Senates

, Ayyildiz

et al. Characterization of nonalcoholic fatty liver disease unrelated to the metabolic syndrome. Eur J Clin Invest, 2012; 42:411–418.

16.

Vanni

, Bugianesi

, Kotronen

et al.

From the metabolic syndrome to NAFLD or vice versa?

Dig Liver Dis, 2010; 42:320–330.

17.

, Shi

, Liu

et al.

Non-alcoholic fatty liver disease: An early mediator predicting metabolic syndrome in obese children?

World J Gastroenterol, 2011; 17:735–742.

18.

LeRoith

. Gender differences in metabolic disorders. Gend Med, 2009; 6,Suppl 1:1–3.

19.

Ayonrinde

, Olynyk

, Beilin

et al. Gender-specific differences in adipose distribution and adipocytokines influence adolescent nonalcoholic fatty liver disease. Hepatology, 2011; 53:800–809.

20.

Tomiya

. Sex difference in non-alcoholic fatty liver disease and non-alcoholic steatohepatitis. Hepatol Res, 2010; 40:108–110.

21.

Descovich

, Cavallo

, Mannino

et al. The Brisighella survey: A study of dietary habits. Minerva Med, 1983; 74:547–555.

22.

ERICA Project Stuff. The CHD risk map of Europe. The 1^st report of the WHO-ERICA. Project Eur Heart J, 1988; 9:1–35.

23.

Seccareccia

, Lanti

, Menotti

et al. Role of body mass index in the prediction of all cause mortality in over 62,000 men and women. The Italian RIFLE Pooling Project. Risk Factor and Life Expectancy. J Epidemiol Community Health, 1988; 52:20–26.

24.

Menotti

, Lanti

, Puddu

et al. The risk functions incorporated in Riscard 2002: A software for the prediction of cardiovascular risk in the general population based on Italian data. Ital Heart J, 2002; 3:114–121.

25.

Menotti

, Lanti

, Agabiti-Rosei

et al. Riskard 2005. New tools for prediction of cardiovascular disease risk derived from Italian population studies. Nutr Metab Cardiovasc Dis, 2005; 15:426–440.

26.

Ardigò

, Bernini

, Borghi

et al. on behalf of the ADSL project investigators. Advanced diagnostic support in lipidology project: Role for phenotypic and functional evaluation of lipoproteins in dyslipidemias. Clin Lipidol, 2010; 5:329–337.

27.

Cicero

AFG

, Dormi

, D'Addato

et al. on behalf of the Brisighella Heart Study Staff. From risk factor assessment to cardiovascular disease risk and mortality modification: The first 40 years of the Brisighella Heart Study. Clin Lipidol, 2011; 6:269–276.

28.

Lee

, Kim

et al. Hepatic steatosis index: A simple screening tool reflecting nonalcoholic fatty liver disease. Dig Liver Dis, 2010; 42:503–508.

29.

Bedogni

, Kahn

, Bellentani

et al. A simple index of lipid overaccumulation is a good marker of liver steatosis. BMC Gastroenterol, 2010; 10:98.

30.

Kleiner

, Brunt

, Van

et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology, 2005; 41:1313–1321.

31.

Angulo

, Hui

, Marchesini

et al. The NAFLD fibrosis score: A noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatology, 2007; 45:846–854.

32.

Dowman

, Tomlinson

, Newsome

. Systematic review: The diagnosis and staging of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis. Aliment Pharmacol Ther, 2011; 33:525–540.

33.

Kahn

. The “lipid accumulation product” performs better than the body mass index for recognizing cardiovascular risk: A population-based comparison. BMC Cardiovasc Disord, 2005; 5:26.

34.

Kahn

. The lipid accumulation product is better than BMI for identifying diabetes: A population-based comparison. Diabetes Care, 2006; 29:151–153.