Comparison of the relationships of alcoholic and nonalcoholic fatty liver with hypertension, diabetes mellitus, and dyslipidemia

Abstract

We compared the relationships of alcoholic fatty liver and nonalcoholic fatty liver with hypertension, diabetes mellitus, and dyslipidemia. Using a nationwide Japanese survey, we collected data on subjects with biopsy-proven alcoholic fatty liver or nonalcoholic fatty liver. Multiple logistic regression analysis was performed to determine whether alcoholic fatty liver and nonalcoholic fatty liver are associated factors for these diseases. Data on 191 subjects (65, alcoholic fatty liver; 126, nonalcoholic fatty liver) were analyzed. Alcoholic fatty liver (odds ratio, 2.54; 95% confidence interval, 1.06–6.32; p = 0.040), age ≥55 years, and body mass index ≥25 kg/m² were correlated with hypertension, whereas nonalcoholic fatty liver (odds ratio, 2.32; 95% confidence interval, 1.08–5.20; p = 0.035) and serum γ-glutamyl transpeptidase levels ≥75 IU/l were correlated with dyslipidemia. Furthermore, we found that there were biological interactions between alcoholic fatty liver and body mass index ≥25 kg/m² in ≥55-year-old subjects (attributable proportion due to interaction, 0.68; 95% confidence interval, 0.19–1.17), as well as between alcoholic fatty liver and age ≥55 years in subjects with body mass index ≥25 kg/m² (attributable proportion due to interaction, 0.71; 95% confidence interval, 0.24–1.18). Alcoholic fatty liver was more strongly associated with hypertension than nonalcoholic fatty liver and nonalcoholic fatty liver was more strongly associated with dyslipidemia than alcoholic fatty liver. Moreover, alcoholic fatty liver, obesity, and older age may interact to influence hypertension status.

Introduction

Fatty liver disease (FLD) is the most prevalent form of liver disease worldwide.^(1,2) Overnutrition and excessive alcohol consumption are 2 major causes of FLD.⁽³⁾ Overnutrition can induce nonalcoholic fatty liver disease (NAFLD), a spectrum of conditions raging from simple steatosis [or nonalcoholic fatty liver (NAFL)] to nonalcoholic steatohepatitis and cirrhosis.^(1,4) NAFLD is considered the hepatic manifestation of the metabolic syndrome,^(5,6) and many studies have revealed strong relationships between NAFLD and hypertension (HT), type 2 diabetes mellitus (DM), and dyslipidemia (DL).⁽⁷⁾ In contrast, excessive alcohol consumption can lead to alcoholic liver disease (ALD), which includes simple steatosis [or alcoholic fatty liver (AFL)], alcoholic hepatitis, hepatic fibrosis, and cirrhosis.⁽²⁾ Less data are available on the relationship between ALD and HT, DM, and DL than that between NAFLD and such diseases. However, accumulating evidence indicates a positive relationship between excessive alcohol consumption and HT, DM, and DL.^(8–11) These findings indicate that ALD may also be closely linked to these diseases.

In the comprehensive management of FLD, it is important to understand the relationships between FLD and HT, DM, and DL in detail. To the best of our knowledge, no study has analyzed these relationships according to the FLD type. The goal of the present study was to compare AFL and NAFL, the most common FLD types, using data from a nationwide Japanese survey on FLD.

Materials and Methods

A nationwide survey

We conducted a nationwide Japanese survey on the status of FLD between 2009 and 2010 by sending a questionnaire to 894 institutions that employed medical specialists in gastroenterology and hepatology. The questionnaire contained questions regarding how histories were taken to assess alcohol consumption and what values were used as the upper limit of alcohol consumption for the purpose of defining social drinking. We also sent data sheets for subjects with biopsy-proven AFL, NAFL, or nonalcoholic steatohepatitis. The data sheets included details regarding age, gender, anthropometric measurements, blood pressure, liver function tests, data regarding the presence or absence of HT, DM, and DL, and laboratory test values associated with these diseases. Data obtained around the time of liver biopsy were collected. Written informed consent was obtained from each patient at the time of biopsy. This study was performed in accordance with the Declaration of Helsinki.

Subjects

In this study, we analyzed data from subjects with AFL or NAFL. AFL was diagnosed according to the following criteria of the Alcohol and Liver Research Group of the Ministry of Education: alcohol consumption ≥60 g/day for >5 years for men and ≥40 g/day for >5 years for women.⁽¹²⁾ For the diagnosis of NAFL, we adopted ⩽20 g/day as the upper limit of alcohol consumption, a value that is accepted by most researchers.^(5,13)

Criteria for HT, DM, and DL

The diagnosis of HT, DM, and DL was made on the basis of treatments for these diseases or their respective criteria defined below. HT was defined according to the following Japanese Society of Hypertension guidelines for the management of hypertension: a systolic blood pressure ≥140 mmHg or a diastolic blood pressure ≥90 mmHg.⁽¹⁴⁾ DM was defined by the following criteria of the Japan Diabetes Society: fasting blood glucose levels ≥126 mg/dl or random blood glucose levels ≥200 mg/dl.⁽¹⁵⁾ DL was defined as serum low-density-lipoprotein (LDL) cholesterol levels ≥140 mg/dl, serum high-density-lipoprotein (HDL) cholesterol levels <40 mg/dl, or serum triglyceride levels ≥150 mg/dl, according to the criteria of the Japan Atherosclerosis Society.⁽¹⁶⁾ Serum LDL cholesterol levels were calculated using the Friedewald equation (LDL cholesterol = total cholesterol − HDL cholesterol − triglycerides/5) for subjects with serum triglyceride levels <400 mg/dl.⁽¹⁷⁾

Statistical analysis

Data are expressed as medians (ranges) or percentages. The chi-square test or Fisher’s exact probability test were used to compare categorical variables, and the Mann–Whitney U test was used to compare continuous variables. A multiple logistic regression analysis was performed to determine whether the FLD types were associated factors for HT, DM, DL or combinations thereof (HT + DM, HT + DL, DM + DL, HT + DM + DL, and all combinations of ≥2 of the 3 diseases). The following potential confounding variables were included in the analysis: age (≥55 years, <55 years), gender (male, female), body mass index (BMI) (≥25 kg/m², <25 kg/m²), serum aspartate aminotransferase (AST) levels (≥40 IU/L, <40 IU/L), and serum γ-glutamyl transpeptidase (γ-GTP) levels (≥75 IU/L, <75 IU/L). BMI ≥25 kg/m² is defined as obesity in Japan.⁽¹⁸⁾ A p value <0.05 was considered statistically significant. If the FLD types and other variables were simultaneously identified as associated factors, stratified and biological interaction analyses were conducted. Three indices were employed to assess biological interaction: the relative excess risk due to interaction (RERI), the attributable proportion due to interaction (AP), and the synergy index (S).^(19,20) Methods for calculating the indices and their 95% confidence intervals (CI) are described by Andersson et al.⁽²¹⁾ RERI = 0, AP = 0, or S = 1 indicated an additive interaction; RERI >0, AP >0, or S >1 indicated a synergistic interaction; and RERI <0, AP <0, or S <1 indicated an antagonistic interaction.⁽²²⁾ Analyses were performed using STATA ver 11.1 (STATA Corp., College Station, TX).

Results

Answers to the questionnaire were obtained from 101 (11.3%) of the 894 institutions and data on FLD were obtained from 66 hospitals (7.4%). The numbers of patients with FLDs were as follows: AFL, 71; NAFL, 131; nonalcoholic steatohepatitis, 494. Of the 202 subjects with AFL or NAFL, 6 with AFL and 5 with NAFL were excluded because of a lack of anthropometric data or information on the presence/absence of HT, DM, and DL. Thus, this study was conducted using data for 191 subjects (65, AFL; 126, NAFL).

Table 1 shows the baseline characteristics of the subjects. The female-to-male ratio, BMI, and diastolic blood pressure were lower in subjects with AFL than in those with NAFL. Laboratory tests revealed that levels of serum AST, γ-GTP, and fasting blood glucose were higher in subjects with AFL than in those with NAFL, whereas levels of serum total cholesterol and LDL cholesterol were lower in subjects with AFL than in those with NAFL. There were no significant differences in prevalence of HT, DM, DL, HT + DM, HT + DL, DM + DL, HT + DM + DL, and any combination of ≥2 of the 3 diseases between subject groups.

Table 1 Baseline characteristics of the subjects

	Total cohort (n = 191)	AFL (n = 65)	NAFL (n = 126)	p value*
Age, years	54 (15–85)	56 (23–80)	53.5 (15–85)	0.541
Gender, male/female	115/76	50/15	65/61	0.0007
BMI, kg/m2	24.9 (13.2–61.3)	24.4 (18.0–35.3)	25.4 (13.2–61.3)	0.026
Systolic blood pressure, mmHg	124 (84–186)	123 (100–186)	125 (84–168)	0.727
Diastolic blood pressure, mmHg	76 (46–114)	74 (58–114)	78 (46–107)	0.010
AST, IU/L	41 (15–675)	61 (17–675)	35 (15–310)	<0.0001
ALT, IU/L	49 (12–1123)	49 (12–1123)	48.5 (13–377)	0.827
γ-GTP, IU/L	72 (10–3028)	156 (24–3028)	50 (10–646)	<0.0001
Fasting blood glucose, mg/dL	103 (67–310)	111.5 (70–176)	100 (67–310)	0.005
Hemoglobin A1c, %	5.8 (3.7–10.6)	5.9 (3.7–9.1)	5.8 (4.4–10.6)	0.450
Total cholesterol, mg/dL	192 (37–454)	166 (37–454)	201 (88–349)	<0.0001
LDL cholesterol, md/dL†	114 (5–246)	89 (5–210)	119 (51–246)	<0.0001
HDL cholesterol, mg/dL	49 (3–131)	47 (3–131)	50 (20–129)	0.182
Triglycerides, mg/dL	118 (21.5–879)	110 (25–879)	120 (21.5–407)	0.731
HT, n (%)	60 (31.4)	25 (38.5)	35 (27.8)	0.132
untreated, n (%)	18 (30.0)	10 (40.0)	8 (22.9)
under treatment, n (%)	42 (70.0)	15 (60.0)	27 (77.1)
DM, n (%)	47 (24.6)	19 (29.2)	28 (22.2)	0.287
untreated, n (%)	18 (38.3)	10 (52.6)	8 (28.6)
under treatment, n (%)	29 (61.7)	9 (47.4)	20 (71.4)
DL, n (%)	71 (37.2)	21 (32.3)	50 (39.7)	0.318
untreated, n (%)	43 (60.6)	15 (71.4)	28 (56.0)
under treatment, n (%)	28 (39.4)	6 (28.6)	22 (44.0)
HT + DM, n (%)	26 (13.6)	12 (18.5)	14 (11.1)	0.184
HT + DL, n (%)	34 (17.8)	12 (18.5)	22 (17.5)	0.864
DM + DL, n (%)	26 (13.6)	6 (9.2)	20 (15.9)	0.267
HT + DM + DL, n (%)	16 (8.4)	5 (7.7)	11 (8.7)	1.000
≥2 of the 3 diseases, n (%)	53 (27.7)	20 (30.8)	33 (26.2)	0.503

AFL, alcoholic fatty liver; NAFL, nonalcoholic fatty liver; BMI, body mass index; AST, aspartate aminotransferase; ALT, alanine aminotransferase; γ-GTP, γ-glutamyl transpeptidase; HT, hypertension; DM, diabetes mellitus; DL, dyslipidemia; LDL, low-density lipoprotein; HDL, high-density lipoprotein. *AFL vs NAFL. Chi-square test or Fisher’s exact probability test for categorical variables, Mann–Whitney U test for continuous variables. ^†The Friedewald equation was used. Data excluded 4 subjects with AFL and 1 with NALF whose serum triglyceride levels were ≥400 mg/dL.

Table 2 lists factors associated with HT, DM, or DL for the entire cohort. Regarding FLD types, AFL was an associated factor for HT whereas NAFL was one for DL. Age ≥55 years was identified as a significant factor for HT, DM, and all combinations of the diseases. BMI ≥25 kg/m² was significantly associated with HT and HT + DL. Serum γ-GTP ≥75 IU/L was another factor associated with DL.

Table 2 Associated factors for HT, DM, DL or their combinations

Disease	Associated factors	p value	Adjusted odds ratio*	95% Confidence interval
HT	AFL	0.040	2.54	1.06–6.32
	Age ≥55 years	<0.0001	6.64	3.24–14.49
	BMI ≥25 kg/m2	0.012	2.49	1.23–5.17
DM	Age ≥55 years	<0.0001	5.46	2.55–12.62
DL	NAFL	0.035	2.32	1.08–5.20
	γ-GTP ≥75 IU/L	0.004	2.85	1.42–5.90
HT + DM	Age ≥55 years	0.0006	7.17	2.55–25.78
HT + DL	Age ≥55 years	0.001	4.20	1.81–10.72
	BMI ≥25 kg/m2	0.021	2.61	1.18–6.06
DM + DL	Age ≥55 years	0.002	5.14	1.95–15.64
HT + DM + DL	Age ≥55 years	0.006	8.51	2.22–56.32
≥2 of the 3 diseases	Age ≥55 years	0.0001	4.42	2.19–9.41

HT, hypertension; DM, diabetes mellitus; DL, dyslipidemia; AFL, alcoholic fatty liver; BMI, body mass index; NAFL, nonalcoholic fatty liver; γ-GTP, γ-glutamyl transpeptidase. *A multiple logistic regression analysis was performed on the basis of types of fatty liver disease, age, gender, BMI, and serum levels of aspartate aminotransferase and γ-GTP.

Stratified analysis was performed with regard to the 3 associated factors (AFL, age ≥55 years, and BMI ≥25 kg/m²) for HT. First, the entire cohort was divided into 2 subgroups according to age (≥55 years, <55 years), and stratified analysis by FLD type and BMI was performed. Among subjects aged ≥55 years (n = 94), the prevalence of HT was the highest in subjects with AFL + BMI ≥25 kg/m² (Fig. 1A). After adjusting for other variables, AFL + BMI ≥25 kg/m² showed significantly higher odds ratio (OR) for HT compared with NAFL + BMI <25 kg/m² (Table 3). In interaction analysis between AFL and BMI ≥25 kg/m², AP (0.68, 95% CI, 0.19–1.17) was significant, whereas RERI and S were not. Among subjects aged <55 years (n = 97), the prevalence of HT was the highest in subjects with AFL + BMI <25 kg/m² (Fig. 1B). There were no significant differences between the relationship of each stratified group with HT (Table 3); moreover, neither RERI, AP, nor S was significant as per the interaction analysis.

Fig. 1

Comparison of the prevalence of hypertension among subjects stratified by types of fatty liver disease and body mass index in each age subgroup. (A) Subjects aged ≥55 years. p = 0.082 (chi-square test). (B) Subjects aged <55 years. p = 0.284 (chi-square test). AFL, alcoholic fatty liver; NAFL, nonalcoholic fatty liver; BMI, body mass index ■, presence of hypertension; □, absence of hypertension.

Table 3 Stratified and biological interaction analyses for hypertension in age subgroups

Subgroup	Stratification	p value	Adjusted odds ratio*	95% Confidence interval	Measures of interaction	Estimate	95% Confidence interval
Age ≥55 years (n = 94)	NAFL + BMI <25 kg/m2		1.00
	NAFL + BMI ≥25 kg/m2	0.138	2.29	0.77–6.83	RERI	7.48	–9.59–24.56
	AFL + BMI <25 kg/m2	0.294	2.23	0.50–9.96	AP	0.68	0.19–1.17
	AFL + BMI ≥25 kg/m2	0.008	11.00	1.90–63.86	S	3.97	0.62–25.56
Age <55 years (n = 97)	NAFL + BMI <25 kg/m2		1.00
	NAFL + BMI ≥25 kg/m2	0.127	5.46	0.41–72.11	RERI	–6.50	–27.33–14.33
	AFL + BMI <25 kg/m2	0.098	7.26	0.69–76.09	AP	–1.25	–5.50–3.01
	AFL + BMI ≥25 kg/m2	0.216	5.22	0.38–71.32	S	0.39	0.04–3.80

NAFL, nonalcoholic fatty liver; BMI, body mass index; AFL, alcoholic fatty liver; RERI, relative excess risk due to interaction; AP, attributable proportion due to interaction; S, synergy index. *Odds ratios for hypertension were calculated after adjustment for gender and serum levels of aspartate aminotransferase and γ-glutamyl transpeptidase.

We next divided the entire cohort into 2 subgroups according to BMI (≥25 kg/m², <25 kg/m²), and performed stratified analysis by FLD type and age. Among subjects with BMI ≥25 kg/m² (n = 93), the prevalence of HT was the highest in subjects with AFL + age ≥55 years (Fig. 2A). After adjusting for other variables, AFL + age ≥55 years and NAFL + age ≥55 years showed significantly higher ORs for HT compared with NAFL + age <55 years (Table 4). Regarding interaction analysis between AFL and age ≥55 years, AP (0.71, 95% CI, 0.24–1.18) was significant, whereas RERI and S were not. Among subjects with BMI <25 kg/m² (n = 98), the prevalence of HT was the highest in subjects with AFL + age ≥55 years (Fig. 2B). After adjustment for other variables, AFL + age ≥55 years and NAFL + age ≥55 years showed significantly higher ORs for HT compared with NAFL + age <55 years (Table 4). Neither RERI, AP, nor S was significant as per the interaction analysis.

Fig. 2

Comparison of the prevalence of hypertension among subjects stratified by types of fatty liver disease and age in each body mass index (BMI) subgroup. (A) Subjects with BMI ≥25 kg/m². p<0.0001 (chi-square test). (B) Subjects with BMI <25 kg/m². p = 0.006 (chi-square test). AFL, alcoholic fatty liver; NAFL, nonalcoholic fatty liver; ■, presence of hypertension; □, absence of hypertension.

Table 4 Stratified and biological interaction analyses for hypertension in BMI subgroups

Subgroup	Stratification	p value	Adjusted odds ratio*	95% Confidence interval	Measures of interaction	Estimate	95% Confidence interval
BMI ≥25 kg/m2 (n = 93)	NAFL + Age <55 years		1.00
	NAFL + Age ≥55 years	0.004	5.83	1.79–19.05	RERI	14.43	–16.50–45.35
	AFL + Age <55 years	0.891	1.14	0.17–7.55	AP	0.71	0.24–1.18
	AFL + Age ≥55 years	0.0003	20.40	3.96–104.98	S	3.90	0.66–23.11
BMI <25 kg/m2 (n = 98)	NAFL + Age <55 years		1.00
	NAFL + Age ≥55 years	0.011	16.65	1.90–146.17	RERI	–1.21	–34.15–31.74
	AFL + Age <55 years	0.059	9.61	0.92–100.97	AP	–0.05	–1.46–1.36
	AFL + Age ≥55 years	0.006	24.05	2.44–237.10	S	0.95	0.24–3.84

BMI, body mass index; NAFL, nonalcoholic fatty liver; AFL, alcoholic fatty liver; RERI, relative excess risk due to interaction; AP, attributable proportion due to interaction; S, synergy index. *Odds ratios for hypertension were calculated after adjustment for gender and serum levels of aspartate aminotransferase and γ-glutamyl transpeptidase.

The subjects were divided according to the FLD type and serum γ-GTP levels. The prevalence of DL was the highest in subjects with NAFL + γ-GTP ≥75 IU/L (Fig. 3). After adjusting for other variables, NAFL + γ-GTP ≥75 IU/L showed significantly higher ORs for DL than AFL + γ-GTP <75 IU/L (Table 5). A significant interaction between NAFL and γ-GTP ≥75 IU/L was not detected.

Fig. 3

Comparison of the prevalence of dyslipidemia among subjects stratified by types of fatty liver disease and serum γ-glutamyl transpeptidase level. p = 0.027 (chi-square test). AFL, alcoholic fatty liver; NAFL, nonalcoholic fatty liver; γ-GTP, γ-glutamyl transpeptidase; ■, presence of hypertension; □, absence of hypertension.

Table 5 Stratified and biological interaction analyses for dyslipidemia

Stratification	p value	Adjusted odds ratio*	95% Confidence interval	Measures of interaction	Estimate	95% Confidence interval
AFL + γ-GTP <75 IU/l		1.00
AFL + γ-GTP ≥75 IU/l	0.298	2.13	0.51–8.91	RERI	2.30	–1.64–6.23
NAFL + γ-GTP <75 IU/l	0.482	1.65	0.41–6.61	AP	0.45	–0.07–0.97
NAFL + γ-GTP ≥75 IU/l	0.028	5.08	1.19–21.60	S	2.29	0.46–11.40

AFL, alcoholic fatty liver; γ-GTP, γ-glutamyl transpeptidase; NAFL, nonalcoholic fatty liver; RERI, relative excess risk due to interaction; AP, attributable proportion due to interaction; S, synergy index. *Odds ratios for dyslipidemia were calculated after adjustment for age, gender, body mass index, and serum aspartate aminotransferase levels.

Discussion

To the best of our knowledge, this is the first study to analyze the relationships between FLD and HT, DM, and DL according to the FLD type. Our results show that the FLD type influences the relationship between FLD and HT or DL. Thus, AFL was more strongly associated with HT than NAFL and NAFL was more strongly associated with DL than AFL. In contrast, the relationship between FLD and DM or the combinations of HT, DM, and DL were found not to be influenced by the FLD type.

Intensive studies have established excessive alcohol consumption as a risk factor for HT.⁽⁹⁾ Because hepatic steatosis occurs in almost all subjects who consume alcohol excessively,⁽²³⁾ the close relationship between AFL and HT was expected. However, a full understanding of the influence of the FLD type on the relationship between FLD and HT is still lacking. Studies have revealed various mechanisms by which excessive alcohol consumption induces HT, including increased sympathetic nervous system activity and stimulation of the renin–angiotensin–aldosterone system.⁽⁹⁾ These mechanisms also have been reported to be involved in the metabolic syndrome, which is usually accompanied by NAFLD.^(24,25) Insulin resistance, a key factor in the development of the metabolic syndrome,⁽²⁶⁾ is another mechanism in the pathogenesis of HT.⁽²⁷⁾ Recent studies have found that excessive alcohol consumption does not significantly increase insulin resistance.⁽²⁸⁾ Nevertheless, our study demonstrates that AFL is the FLD type more closely linked to HT. Potential differences in mechanisms of HT between the FLD types might influence planning therapeutic strategies for HT in subjects with such FLD types.

This study identified obesity and older age as other associated factors for HT. These factors are established risk factors for HT.^(29,30) In stratified analysis, the combination of AFL and obesity in older subjects and that of AFL and older age in both obese and nonobese subjects showed a significant increase in the ORs for HT. In interaction analysis, the results differed according to the indices for biological interaction. According to the interaction analysis for AFL and obesity in older subjects, AP was significant, whereas RERI and S were not. The relationship between AFL and older age in obese subjects followed this same pattern. However, a study on interaction analysis published in 2006 demonstrates that AP is the most robust measure in a logistic regression model.⁽³¹⁾ Hence our results could indicate that AFL, obesity, and older age interact to influence hypertension status.

Cross-sectional studies have suggested an interactive influence of excessive alcohol consumption and obesity on HT.^(32,33) Moreover, in overweight men, combined intervention involving restricted alcohol and food consumption leads to decreases in blood pressure more effectively than either intervention alone.⁽³⁴⁾ We were unable to find any published studies examining the interaction between excessive alcohol consumption and older age in relation to HT. On the basis of the theory of biological interaction,⁽³⁵⁾ the interactions found in our present study may indicate the presence of at least a pathway toward HT in which AFL, obesity, and older age, are all involved. However, future prospective studies will be necessary to confirm these interactions.

We show here that NAFL and increased serum γ-GTP levels are associated factors for DL, and their combination is most strongly associated with DL. We could not confirm the DL types with which these factors were associated because we did not collect the relevant information. Generally, baseline serum LDL cholesterol levels were higher in subjects with NAFL than in those with AFL, although the results were calculated using data from untreated as well as treated subjects. This finding is consistent with the results in large-scale studies investigating the influence of alcohol consumption on serum lipid levels in which serum LDL cholesterol levels were inversely correlated with alcohol consumption.^(36,37) Recent studies of subjects with DM have shown that serum γ-GTP levels are positively associated with DL.⁽³⁸⁾ Furthermore, elevation of serum γ-GTP levels has been identified as a predictor for cardiovascular diseases⁽³⁹⁾ as well as a marker of metabolic syndrome.⁽⁴⁰⁾ The complexes that form between γ-GTP forms and LDL lipoprotein facilitate the evolution of atherosclerotic plaques toward instability and rupture.⁽⁴¹⁾ Given these findings, the magnitude of risk for cardiovascular diseases in subjects with NAFL with elevated serum γ-GTP levels should be investigated.

There are some limitations to this study. First, because of its cross-sectional design, this study could not determine causality between HT, DM, and DL and associated factors. Second, the number of subjects was relatively small, constraining statistical power. Third, data on FLD were obtained from only a limited number of institutions in Japan, which might limit generalizability of the findings. Fourth, this study lacked data on smoking, a potential confounder in particular for HT.⁽⁴²⁾ Since a close link of excessive alcohol consumption to smoking has been reported,⁽⁴³⁾ it is possible that in our cohort, the proportion of smokers was higher in subjects with AFL than in those with NAFL. A large-scale study, however, has demonstrated that smoking has a smaller impact on elevation of blood pressure than excessive alcohol consumption in men, no such effect was seen in women.⁽⁴⁴⁾ Therefore, although our results should be interpreted with caution, we feel confident in concluding that they would not have changed significantly if smoking had been included as a variable.

The present study demonstrates that the relationships between FLD and HT, DM, and DL partly depend on the FLD type. We believe that these findings may be helpful in managing subjects with FLD. Future studies are needed to confirm our results and clarify mechanisms responsible for the development of HT in which AFL, obesity, and older age play a role.

Conflict of Interest

We have no financial disclosure or conflict of interest to report.

Acknowledgments

This work was supported in part by a grant from the Health, Labour and Welfare Ministry of Japan and by grant SR2012-04 for research from Kanazawa Medical University, Japan. We are extremely grateful to the following institutions for providing detailed data on subjects with FLD: Sapporo Medical University Hospital, National Hospital Organization Hokkaido Cancer Center, Hokkaido P.W.F.A.C. Sapporo-Kosei General Hospital, Kushiro Rosai Hospital, Sapporo Social Insurance General Hospital, Hirosaki University School of Medicine and Hospital, Iwate Medical University Hospital, Tohoku University Hospital, Shiogama City Hospital, Yamagata University Hospital, Fukushima Medical University Hospital, Tsukuba University Hospital, Ashikaga Red Cross Hospital, Gunma University Hospital, Maebashi Red Cross Hospital, National Hospital Organization Takasaki General Medical Center, National Defense Medical College Hospital, Saitama Medical University Hospital, Dokkyo Medical University Koshigaya Hospital, Chiba University Hospital, Teikyo University Hospital, Tokyo Women’s Medical University Hospital, Toho University Omori Medical Center, Toshiba General Hospital, Toho University Ohashi Medical Center, Showa University Fujigaoka Hospital, St. Marianna University School of Medicine Hospital, Tokai University Hospital, Nippon Medical University Musashi Kosugi Hospital, Kanto Rosai Hospital, Hiratsuka City Hospital, Fujisawa Shounandai Hospital, University of Yamanashi Hospital, Shinshu University Hospital, Kanazawa University Hospital, National Hospital Organization Kanazawa Medical Center, Ishikawa-ken Saiseikai Kanazawa Hospital, University of Fukui Hospital, Iwata City Hospital, Aichi Medical University Hospital, Mie University Hospital, Shiga University of Medical Science Hospital, Fukuchiyama City Hospital, Kyoto Prefectural Yosanoumi Hospital, Osaka University Hospital, Osaka City University Hospital, Kansai Medical University Takii Hospital, Hyogo-Chuo National Hospital, Yamato Takada Municipal Hospital, Nara Prefectural Gojo Hospital, Okayama Saiseikai General Hospital, Hiroshima University Hospital, Tsuchiya General Hospital, Shakaihoken Shimonoseki Kosei Hospital, Ehime University Hospital, Matsuyama Shimin Hospital, Kochi Medical School Hospital, Kurume University Hospital, Asakura Medical Association Hospital, Saiseikai Futsukaichi Hospital, Nagasaki University Hospital, Nagasaki Municipal Hospital, Oita University Hospital, Okinawa Prefectural Chubu Hospital, and Nakagami Hospital. We also thank Minemi Shibayama and Kiri Nakashima for their excellent research assistance.

References

• 1   Lewis  JR,  Mohanty  SR. Nonalcoholic fatty liver disease: a review and update. Dig Dis Sci 2010; 55: 560–578.
• 2   Gao  B,  Bataller  R. Alcoholic liver disease: pathogenesis and new therapeutic targets. Gastroenterology 2011; 141: 1572–1585.
• 3   Mills  SJ,  Harrison  SA. Comparison of the natural history of alcoholic and nonalcoholic fatty liver disease. Curr Gastroenterol Rep 2005; 7: 32–36.
• 4   Vernon  G,  Baranova  A,  Younossi  ZM. Systematic review: the epidemiology and natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther 2011; 34: 274–285.
• 5   Hamaguchi  M,  Kojima  T,  Takeda  N, et al. The metabolic syndrome as a predictor of nonalcoholic fatty liver disease. Ann Intern Med 2005; 143: 722–728.
• 6   Moore  JB. Non-alcoholic fatty liver disease: the hepatic consequence of obesity and the metabolic syndrome. Proc Nutr Soc 2010; 69: 211–220.
• 7   Okanoue  T,  Umemura  A,  Yasui  K,  Itoh  Y. Nonalcoholic fatty liver disease and nonalcoholic steatohepatitis in Japan. J Gastroenterol Hepatol 2011; 26 (Suppl 1): 153–162.
• 8   Nakanishi  N,  Yoshida  H,  Nakamura  K,  Suzuki  K,  Tatara  K. Alcohol consumption and risk for hypertension in middle-aged Japanese men. J Hypertens 2001; 19: 851–855.
• 9   Miller  PM,  Anton  RF,  Egan  BM,  Basile  J,  Nguyen  SA. Excessive alcohol consumption and hypertension: clinical implications of current research. J Clin Hypertens (Greenwich) 2005; 7: 346–351.
• 10   Seike  N,  Noda  M,  Kadowaki  T. Alcohol consumption and risk of type 2 diabetes mellitus in Japanese: a systematic review. Asia Pac J Clin Nutr 2008; 17: 545–551.
• 11   Van de Wiel  A. The effect of alcohol on postprandial and fasting triglycerides. Int J Vasc Med 2012; 2012: 862504.
• 12   Takada  A,  Tsutsumi  M. Diagnostic criteria for alcoholic liver disease. Int Hepatol Comm 1995; 3: 63–69.
• 13   Chitturi  S,  Farrell  GC,  Hashimoto  E,  Saibara  T,  Lau  GK,  Sollano  JD. Non-alcoholic fatty liver disease in the Asia-Pacific region: definitions and overview of proposed guidelines. J Gastroenterol Hepatol 2007; 22: 778–787.
• 14   Ogihara  T,  Kikuchi  K,  Matsuoka  H, et al. The Japanese society of hypertension guidelines for the management of hypertension (JSH 2009). Hypertens Res 2009; 32: 3–107.
• 15   Kuzuya  T,  Nakagawa  S,  Satoh  J, et al. Report of the Committee of Japan Diabetes Society on the classification and diagnostic criteria of diabetes mellitus. J Japan Diab Soc 1999; 42: 385–404(in Japanese with English abstract).
• 16   Teramoto  T,  Sasaki  J,  Ueshima  H, et al. Diagnostic criteria for dyslipidemia. Executive summary of Japan Atherosclerosis Society (JAS) guideline for diagnosis and prevention of atherosclerotic cardiovascular diseases for Japanese. J Atheroscler Thromb 2007; 14: 155–158.
• 17   Nauck  M,  Warnick  GR,  Rifai  N. Methods for measurement of LDL-cholesterol: a critical assessment of direct measurement by homogeneous assays versus calculation. Clin Chem 2002; 48: 236–254.
• 18   Matsuzawa  Y,  Nakamura  T,  Takahashi  M, et al. New criteria for ‘obesity disease’ in Japan. Circ J 2002; 66: 987–992.
• 19   Rothman  KJ. The estimation of synergy or antagonism. Am J Epidemiol 1976; 103: 506–511.
• 20   Hosmer  DW,  Lemeshow  S. Confidence interval estimation of interaction. Epidemiology 1992; 3: 452–456.
• 21   Andersson  T,  Alfredsson  L,  Källberg  H,  Zdravkovic  S,  Ahlbom  A. Calculating measures of biological interaction. Eur J Epidemiol 2005; 20: 575–579.
• 22   Knol  MJ,  VanderWeele  TJ,  Groenwold  RH,  Klungel  OH,  Rovers  MM,  Grobbee  DE. Estimating measures of interaction on an additive scale for preventive exposures. Eur J Epidemiol 2011; 26: 433–438.
• 23   Lieber  CS. Alcoholic fatty liver: its pathogenesis and mechanism of progression to inflammation and fibrosis. Alcohol 2004; 34: 9–19.
• 24   Mancia  G,  Bousquet  P,  Elghozi  JL, et al. The sympathetic nervous system and the metabolic syndrome. J Hypertens 2007; 25: 909–920.
• 25   Wang  CH,  Li  F,  Takahashi  N. The renin angiotensin system and the metabolic syndrome. Open Hypertens J 2010; 3: 1–13.
• 26   Cornier  MA,  Dabelea  D,  Hernandez  TL, et al. The metabolic syndrome. Endocr Rev 2008; 29: 777–822.
• 27   Reaven  GM. Relationships among insulin resistance, type 2 diabetes, essential hypertension, and cardiovascular disease: similarities and differences. J Clin Hypertens (Greenwich) 2011; 13: 238–243.
• 28   Clerc  O,  Nanchen  D,  Cornuz  J, et al. Alcohol drinking, the metabolic syndrome and diabetes in a population with high mean alcohol consumption. Diabet Med 2010; 27: 1241–1249.
• 29   Rahmouni  K,  Correia  ML,  Haynes  WG,  Mark  AL. Obesity-associated hypertension: new insights into mechanisms. Hypertension 2005; 45: 9–14.
• 30   Mateos-Cáceres  PJ,  Zamorano-León  JJ,  Rodríguez-Sierra  P,  Macaya  C,  López-Farré  AJ. New and old mechanisms associated with hypertension in the elderly. Int J Hypertens 2012; 2012: 150107.
• 31   Kalilani  L,  Atashili  J. Measuring additive interaction using odds ratios. Epidemiol Perspect Innov 2006; 3: 5.
• 32   Li  Y,  Wang  JG,  Gao  PJ, et al. Interaction between body mass index and alcohol intake in relation to blood pressure in HAN and SHE Chinese. Am J Hypertens 2006; 19: 448–453.
• 33   Gu  M,  Qi  Y,  Li  M,  Niu  W. Association of body mass index and alcohol intake with hypertension subtypes among HAN Chinese. Clin Exp Hypertens 2011; 33: 518–524.
• 34   Puddey  IB,  Parker  M,  Beilin  LJ,  Vandongen  R,  Masarei  JR. Effects of alcohol and caloric restrictions on blood pressure and serum lipids in overweight men. Hypertension 1992; 20: 533–541.
• 35   Ahlbom  A,  Alfredsson  L. Interaction: a word with two meanings creates confusion. Eur J Epidemiol 2005; 20: 563–564.
• 36   Nakanishi  N,  Yoshida  H,  Nakamura  K,  Kawashimo  H,  Tatara  K. Influence of alcohol intake on risk for increased low-density lipoprotein cholesterol in middle-aged Japanese men. Alcohol Clin Exp Res 2001; 25: 1046–1050.
• 37   Kato  I,  Kiyohara  Y,  Kubo  M, et al. Insulin-mediated effects of alcohol intake on serum lipid levels in a general population: the Hisayama Study. J Clin Epidemiol 2003; 56: 196–204.
• 38   Zoppini  G,  Targher  G,  Trombetta  M,  Lippi  G,  Muggeo  M. Relationship of serum gamma-glutamyltransferase to atherogenic dyslipidemia and glycemic control in type 2 diabetes. Obesity (Silver Spring) 2009; 17: 370–374.
• 39   Turgut  O,  Yilmaz  A,  Yalta  K,  Karadas  F,  Birhan Yilmaz  M. γ-Glutamyltransferase is a promising biomarker for cardiovascular risk. Med Hypotheses 2006; 67: 1060–1064.
• 40   Ryu  S,  Chang  Y,  Woo  HY, et al. Longitudinal increase in gamma-glutamyltransferase within the reference interval predicts metabolic syndrome in middle-aged Korean men. Metabolism 2010; 59: 683–689.
• 41   Turgut  O,  Tandogan  I. Gamma-glutamyltransferase to determine cardiovascular risk: shifting the paradigm forward. J Atheroscler Thromb 2011; 18: 177–181.
• 42   Grassi  G,  Seravalle  G,  Calhoun  DA, et al. Mechanisms responsible for sympathetic activation by cigarette smoking in humans. Circulation 1994; 90: 248–253.
• 43   Margetts  BM,  Jackson  AA. Interactions between people’s diet and their smoking habits: the dietary and nutritional survey of British adults. BMJ 1993; 307: 1381–1384.
• 44   Primatesta  P,  Falaschetti  E,  Gupta  S,  Marmot  MG,  Poulter  NR. Association between smoking and blood pressure: evidence from the health survey for England. Hypertension 2001; 37: 187–193.