Moderate alcohol consumption is associated with advanced fibrosis in non-alcoholic fatty liver disease and shows a synergistic effect with type 2 diabetes mellitus

(1)

Moderate alcohol consumption is associated with advanced

ﬁbrosis in

non-alcoholic fatty liver disease and shows a synergistic effect with type

2 diabetes mellitus

Julia Blomdahl, Patrik Nasr, Mattias Ekstedt

1

, Stergios Kechagias

⁎

,1

Department of Gastroenterology and Hepatology, Department of Health, Medicine, and Caring Sciences, Linköping University, SE-581 83 Linköping, Sweden

a b s t r a c t

a r t i c l e i n f o

Article history: Received 16 September 2020 Accepted 19 November 2020 Keywords:

Non-alcoholic fatty liver disease Alcohol drinking

Type 2 diabetes mellitus Phosphatidylethanol

Background: Non-alcoholic fatty liver disease (NAFLD) is the most common chronic liver disease worldwide. Whether moderate alcohol consumption plays a role for progression of NAFLD is disputed. Moreover, it is not known which tool is ideal for assessment of alcohol consumption in NAFLD. This study aimed to evaluate if moderate alcohol consumption assessed with different methods, including the biological marker phosphatidylethanol (PEth), is associated with advancedﬁbrosis in NAFLD.

Methods: We conducted a cross-sectional study of patients with biopsy-proven NAFLD. All participants were clin-ically evaluated with medical history, blood tests, and anthropometric measurements. Alcohol consumption was assessed using PEth in blood, the questionnaire AUDIT-C, and clinical interview.

Findings: 86 patients were included of which 17% had advancedfibrosis. All participants reported alcohol con-sumption < 140 g/week. Average weekly alcohol concon-sumption was higher in the group with advancedfibrosis. Moderate alcohol consumption, independently of the method of assessment, was associated with increased probability of advancedfibrosis (adjusted OR 5.5–9.7, 95% CI 1.05–69.6). Patients with type 2 diabetes mellitus (T2DM) consuming moderate amounts of alcohol had a significantly higher rate of advanced fibrosis compared with those consuming low amounts (50.0–60.0% vs. 3.3–21.6%, p < 0.05).

Conclusions: Moderate alcohol consumption, irrespective of assessment method (clinical interview, AUDIT-C, and PEth), was associated with advancedfibrosis. PEth in blood ≥ 50 ng/mL may be a biological marker indicating in-creased risk for advancedfibrosis in NAFLD. Patients with T2DM consuming moderate amounts of alcohol had the highest risk of advancedfibrosis, indicating a synergistic effect of insulin resistance and alcohol on the histopath-ological progression of NAFLD.

1. Introduction

According to European guidelines, weekly alcohol consumption in non-alcoholic fatty liver disease (NAFLD) is set at a maximum of 30 g/ day for males and 20 g/day for females [1]. However, guidelines do not state how the assessment of alcohol consumption should be performed. Commonly, clinical interview and/or information from a

self-reported questionnaire are used. One of the most widespread and validated tools for excluding excessive alcohol consumption in NAFLD is AUDIT [2]. A shorter version of AUDIT (AUDIT-C), consisting of only three questions, has been developed and more largely used [3]. How-ever, people with alcohol misuse will often inaccurately report that they do not have a problem, creating a need for more objective methods to investigate a person's drinking habits.

Analysis of phosphatidylethanol in blood (PEth) has emerged as a sensitive and speci_{ﬁc method to investigate a person's drinking habits,} particularly for social drinkers and risk drinkers [4]. PEth comprises a group of homologous phospholipids found in cell membranes [5], which are formed in the presence of ethanol. No false-positive PEth-values have been reported [6] and it can be detected in blood between 2 and 4 weeks after ingestion [7].

A potentially important factor for clinical and histopathological pro-gression of NAFLD is the impact of moderate alcohol consumption. There is evidence for beneﬁcial effects of modest alcohol consumption on the risk of metabolic syndrome and insulin resistance [8], which

Abbreviations: NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; AUDIT, Alcohol Use Disorders Identification Test; AUDIT-C, Alcohol Use Disorders Identification Test – Consumption; PEth, phosphatidylethanol; LC-MS/MS, liq-uid chromatography tandem mass spectrometry; LOQ, level of quantification; T2DM, type 2 diabetes mellitus; BMI, body mass index; HOMA-IR, homeostatic model assessment for insulin resistance; ALT, alanine transaminase; AST, aspartate transaminase; INR, inter-national normalised ratio.

⁎ Corresponding author.

E-mail addresses:julia.blomdahl@liu.se(J. Blomdahl),patrik.nasr@liu.se(P. Nasr),

mattias.ekstedt@liu.se(M. Ekstedt),stergios.kechagias@liu.se(S. Kechagias).

1

Senior authors contributed equally.

https://doi.org/10.1016/j.metabol.2020.154439

Contents lists available atScienceDirect

Metabolism Clinical and Experimental

j o u r n a l h o m e p a g e :w w w . m e t a b o l i s m j o u r n a l . c o m

(2)

are important components of the NAFLD disease process. However, the impact of moderate alcohol consumption on histopathology and clinical outcome in NAFLD is disputed as studies show con_{ﬂicting results,} stat-ing both improvement and worsenstat-ing of histological parameters in moderate drinkers [9].

The purpose of this study was to evaluate if moderate alcohol con-sumption is associated with the presence of advanced ﬁbrosis in NAFLD patients. Moreover, we evaluated the assessment of alcohol con-sumption in NAFLD with three different methods: PEth, clinical inter-view, and AUDIT-C.

2. Methods

In this cross-sectional study of prospectively enrolled patients with biopsy-proven NAFLD, alcohol consumption was assessed using three different methods, AUDIT-C, clinical interview, and measurement of the direct biomarker PEth in blood. More details on Methods are avail-able in the online-only Supplementary data.

3. Results

3.1. Participant characteristics and advancedﬁbrosis

Baseline characteristics of patients are presented inTable 1. Partici-pants were also divided into two groups based onﬁbrosis stage (F0–2 and F3–4). Patient characteristics in the two groups are presented in

Table 1. Weekly alcohol consumption was higher in the advanced fibro-sis group, both when assessed with AUDIT-C and interview (47.5 and 56.3 g/week compared to 27.6 and 31.2 g/week). Assessment with in-terview attained statistical significance, however, only a trend was seen when assessed with AUDIT-C (p = 0.10). Average PEth was 66.5 ng/mL in the advanced_{fibrosis group and 38.4 ng/mL in participants} with F0–2, but it did not reach statistical significance (p = 0.25).

Correlations between the three methods used for assessment of alcohol consumption were all signiﬁcant. The strongest correlation was seen between consumption assessed with AUDIT-C and clinical interview (r = 0.89, p < 0.01). The correlation coefﬁcient was 0.44 (p < 0.01) between AUDIT-C and PEth and 0.56 (p < 0.01) between interview and PEth.

3.2. Alcohol consumption and diabetes mellitus type 2

To determine if the presence of T2DM was associated with advanced ﬁbrosis, patients were divided into four groups (Fig. 1), 1) low alcohol consumption and no T2DM, 2) low alcohol consumption and T2DM, 3) moderate alcohol consumption and no T2DM, and 4) moderate alco-hol consumption and T2DM. Moderate alcoalco-hol consumption was de-ﬁned using three different cut-offs, A) >66 g/week (AUDIT-C), B) >96 g/week (clinical interview), and C) PEth≥ 50 ng/mL. The different cut-offs were attained from multivariate logistic regression models. Similar results were seen with respect to all three assessments of alcohol con-sumption. The group with moderate alcohol consumption and T2DM

Table 1 Patient characteristics. Total (n = 86) F0–2 (n = 71) F3–4 (n = 15) p-value Age (years) 59.8 (±11.1) 59.6 (±11.7) 60.6 (±7.5) NS Sex (% male/female) 70.9/29.1 70.4/29.6 73.3/26.7 NS BMI (kg/m2 ) 30.2 (±8.6) 30.1 (±9.3) 30.4 (±4.5) NS Hypertension (%) 81.2 82.9 73.3 NS HbA1C (mmol/mol) 43.3 (±11.8) 42.4 (±12.0) 47.3 (±10.0) NS HOMA-IR 4.98 (±4.1) 4.43 (±3.8) 7.64 (±4.7) p = 0.012 Type 2 diabetes mellitus (%) 48.8 43.7 73.3 p = 0.048 Alcohol consumption according to AUDIT-C (g/week) 31.0 (±34.4) 27.6 (±31.9) 47.5 (±42.0) NS Alcohol consumption according to interview (g/week) 35.6 (±40.3) 31.2 (±36.8) 56.3 (±50.7) p = 0.028 PEth (ng/mL) 43.3 (±84.7) 38.4 (±83.1) 66.5 (±91.6) NS Haemoglobin (g/dL) 14.7 (±1.2) 14.6 (±1.2) 15.3 (±1.0) NS Platelets (×109 /L) 232.7 (±57.9) 239.3 (±57.0) 201.1 (±52.8) p = 0.019 Prothrombin (INR) 1.0 (±0.1) 1.0 (±0.1) 1.0 (±0.0) p = 0.003 Albumin (g/L) 41.4 (±3.9) 41.5 (±3.9) 40.8 (±4.0) NS Glucose (mg/dL) 126 (±38) 122 (±38) 142 (±32) NS Ferritin (μg/L) 240.6 (±215.5) 223.7 (±207.5) 320.7 (±241.2) NS ALT (U/L) 67.2 (±21.0) 61.2 (±41.4) 93.6 (±46.8) p = 0.010 AST (U/L) 39.6 (±18.0) 35.9 (±17.8) 58.1 (±21.6) p = 0.000 AST/ALT-ratio 0.71 (±0.4) 0.70 (±0.4) 0.71 (±0.3) NS Steatosis (%) Grade 0 0 0 0 Grade 1 46.5 50.7 26.7 Grade 2 30.2 29.6 33.3 Grade 3 23.3 19.7 40.0 Lobular inﬂammation (%) Grade 0 80.2 91.6 26.7 Grade 1 9.3 5.6 26.7 Grade 2 8.2 1.4 40.0 Grade 3 2.3 1.4 6.6 Ballooning (%) Grade 0 67.4 77.5 20.0 Grade 1 27.9 19.7 66.7 Grade 2 4.7 2.8 13.3 Fibrosis (%) Stage 0 31.4 38.0 – Stage 1 29.0 35.2 – Stage 2 22.1 26.8 – Stage 3 16.3 – 93.3 Stage 4 1.2 – 6.7

Patient characteristics and laboratory values in the total population and divided by the twoﬁbrosis groups. All values are presented as mean (±SD) or % when stated otherwise. Abbreviations: BMI, body mass index; HOMA-IR, homeostatic model assessment for insulin resistance; AUDIT-C, Alcohol Use Disorders Identiﬁcation Test - Consumption; PEth, phosphatidylethanol; INR, international normalised ratio; ALT, alanine aminotransferase; AST, aspartate aminotransferase.

(3)

had signiﬁcantly higher presence of advanced ﬁbrosis compared to the groups with low alcohol consumption (Fig. 1), regardless of the pres-ence of T2DM.

3.3. Multivariate models

Weekly alcohol consumption assessed with AUDIT-C, interview, and PEth was used in a bivariate, and a multivariate logistic regression model, comparing the risk for advanced fibrosis (Supplementary Table 1). In the bivariate unadjusted model, the risk of having advanced fibrosis increased by 1.4–1.6% for each gram of ethanol per week (assessed with interview and AUDIT-C). When adjusted for age, sex, BMI, and diagnosis of T2DM, the risk was even higher, 1.8–1.9%. PEth did not reach statistical significance in the bivariate model. In all three bivariate models, T2DM was significantly associated with increased risk for advancedfibrosis.

In the multivariate logistic regression model, alcohol consumption assessed with AUDIT-C showed a crude OR for advancedﬁbrosis of 6.3 in those consuming >66 g/week as compared to 0–2.99 g/week. After adjusting for age, sex, BMI, and T2DM the OR was even higher (9.7, 95%

CI 1.4–68.9). Alcohol consumption of 3–66 g/week was not associated with significantly increased risk of advanced fibrosis compared with con-sumption of 0–2.99 g/week. Using the same cut-off level in the model of alcohol consumption assessed with clinical interview, statistical signi fi-cance was not reached. Instead, alcohol consumption of more than 96 g/ week was statistically significant compared with consumption of 0–2.99 g/week, with an OR of 7.0 (95% CI 1.1–44.1). Alcohol consumption of 3– 96 g/week was not associated with significantly increased risk of ad-vancedfibrosis compared with consumption of 0–2.99 g/week. After adjusting for age, sex, BMI, and T2DM, the OR of advancedfibrosis in par-ticipants drinking >96 g/week, was even higher (8.5, 95% CI 1.05–69.6). In this model, a PEth-value of≥ 50 ng/mL was significantly associated with advancedfibrosis (OR 3.3, 95% CI 1.02–10.5; adjusted OR 5.5, 95% CI 1.4–22.1). In the adjusted multivariate model for AUDIT-C and PEth, T2DM was significantly associated with advanced fibrosis.

4. Discussion

The mainﬁnding of this study was that moderate alcohol consump-tion, irrespective of assessment method (clinical interview, AUDIT-C,

Alcohol > 66 g/wk, T2DM Alcohol > 66 g/wk, no T2DM Alcohol < 66 g/wk, T2DM Alcohol < 66 g/wk, no T2DM

Percent

100 80 60 40 20 0

Alcohol consumption (AUDIT-C) and T2DM

F3-4 F0-2 5.7 % (n=2) 20.0 % (n=7) 22.2 % (n=2) 57.1 % (n=4) * **

A

Alcohol > 96 g/wk, T2DM Alcohol > 96 g/wk, no T2DM Alcohol < 96 g/wk, T2DM Alcohol < 96 g/wk, no T2DM

Percent

100 80 60 40 20 0

Alcohol consumption (interview) and T2DM

F3-4 F0-2 * ** 7.3 % (n=3) 21.6 % (n=8) 33.3 % (n=1) 60.0 % (n=3)

B

PEth ≥ 50 ng/mL, T2DM PETH ≥ 50 ng/mL, no T2DM PEth ≤ 50 ng/mL, T2DM PEth ≤ 50 ng/mL, no T2DM

Percent

100 80 60 40 20 0

Alcohol consumption (PEth) and T2DM

F3-4 F0-2 3.3 % (n=1) 20.6 % (n=7) 21.4 % (n=3) 50.0 % (n=4) * **

C

Fig. 1. Alcohol consumption and type 2 diabetes mellitus. Advancedfibrosis and alcohol consumption divided by different cut-offs and separated by T2DM. * = p < 0.05; ** = p < 0.01. Abbreviations: AUDIT-C, Alcohol Use Disorders Identification Test – Consumption; T2DM, type 2 diabetes mellitus; PEth, phosphatidylethanol; g/wk, grams/week; F, fibrosis stage.

(4)

and PEth), was associated with advancedﬁbrosis. Moreover, we showed that NAFLD patients with T2DM consuming moderate amounts of alco-hol had the highest risk of advanced_{ﬁbrosis, indicating a synergistic} ef-fect of insulin resistance and alcohol on NAFLD histopathology.

Accurate assessment of alcohol consumption is mandatory for a correct diagnosis of NAFLD. Questionnaires used to assess alcohol consumption have been developed for detecting alcohol abuse, not to evaluate moderate alcohol intake. AUDIT-C, consisting of only three questions, has a good specificity for heavy drinking and dependence [3]. Yet, its effectiveness to diagnose moderate alcohol drinking has never been thoroughly investi-gated. In the present study, we show an excellent correlation (r = 0.89) between AUDIT-C and the more time-consuming clinical interview in indi-viduals with moderate alcohol consumption. Moreover, we show that con-sumption > 66 g/week according to AUDIT-C was associated with advancedfibrosis. The corresponding figure for assessment with clinical interview was slightly higher (>96 g/week). The most probable explana-tion for this discrepancy is the construcexplana-tion of the AUDIT-C quesexplana-tionnaire in which the patient specifies intervals of the number of drinks rather than the actual numbers consumed.

The effect of moderate alcohol consumption on NAFLD seems to be multifaceted. Most studies, including a meta-analysis of over 40,000 indi-viduals [10], indicate that modest alcohol consumption is associated with decreased risk for the development of fatty liver disease [11,12]. In a recent cross-sectional study, Vilar-Gomez et al. [13] examined the association be-tween the gene variant ADH1B*2 and moderate alcohol consumption and histologic severity of NAFLD. The variant gene, ADH1B*2, results in a higher metabolising activity. It was shown that consuming moderate amounts of alcohol and having the variant gene, resulted in less NASH andfibrosis, compared with those without the variant gene. Moreover, consuming moderate amounts of alcohol compared with abstention re-sulted in less NASH andfibrosis, regardless of having the variant gene or not. Studies evaluating ultrasonographical_{findings or levels of serum} ami-notransferases as surrogate markers for NAFLD have mainly shown posi-tive effects of moderate consumption, especially for hepatic steatosis [12,14–18]. However, recent studies [19] have shown thatfibrosis stage, but no other histopathological parameters, determines the future risk of mortality and liver-related morbidity in NAFLD.

In recent years, new non-invasive techniques for diagnosingfibrosis have been developed [20,21], but liver biopsy remains the gold standard for the diagnosis of NAFLD and assessing the severity offibrosis. Hith-erto, 12 studies have assessed the impact of alcohol on histopathology in NAFLD [9]. Robust conclusions cannot be drawn because study design varies and particularly since the results are divergent. In this study, we show an association of moderate alcohol consumption and advanced fi-brosis in NAFLD.

A major strength of our study is the use of an objective marker of al-cohol consumption. Analysis of PEth has hitherto been used only in few NAFLD studies, making it hard to assess its utility. For theﬁrst time, we show that PEth as low as≥50 ng/mL is associated with advanced ﬁbrosis in NAFLD. Our results can be used to aid the interpretation of PEth in NAFLD patients.

5. Conclusions

Although limited by a small sample size, this study indicates that moderate alcohol consumption in NAFLD may aggravatefibrosis and in-crease the risk for future development of end-stage liver disease. NAFLD patients with T2DM consuming moderate amounts of alcohol seem to be at the highest risk for advancedfibrosis. Further prospective studies assessing alcohol consumption using sensitive direct biomarkers, such as PEth, are needed to confirm our results.

CRediT authorship contribution statement

Julia Blomdahl: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation,

Visualization, Writing - original draft, Writing - review & editing. Patrik Nasr: Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Writing - review & editing. Mattias Ekstedt: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing - review & editing. Stergios Kechagias: Conceptualization, Funding acquisition, In-vestigation, Methodology, Project administration, Resources, Supervi-sion, Writing - review & editing.

Declaration of competing interest None.

Acknowledgements

We would like to thank the Clinical Pharmacology Laboratory, Re-gion Östergötland, for the analysis of phosphatidylethanol in blood. We would also like to thank the Medical Research Council of Southeast Sweden for providing funding for our research.

Financial support (funding)

ALF Grants, Region Östergötland, Medical Research Council of South-east Sweden (grant no. 752871). The funding sources had no role in the conduction of the research or the preparation of the article.

Appendix A. Supplementary data

Supplementary data to this article can be found online athttps://doi. org/10.1016/j.metabol.2020.154439.

References

[1] EASL–EASD–EASO Clinical Practice Guidelines for the management of non-alcoholic fatty liver disease. Journal of hepatology. 2016;64(6):1388–402. doi:https://doi. org/10.1016/j.jhep.2015.11.004.

[2]Saunders JB, Aasland OG, Babor TF, De la Fuente JR, Grant M. Development of the Al-cohol Use Disorders Identiﬁcation Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption—II. Addiction. 1993;88: 791–804.

[3]Bush K, Kivlahan D, McDonell M, Fihn S, Bradley K. The AUDIT Alcohol Consumption Questions (AUDIT-C): an effective brief screening test for problem drinking. Ambu-latory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identiﬁ-cation Test. Arch Intern Med. 1998;158:1789–95.

[4] Schrock A, Wurst FM, Thon N, Weinmann W. Assessing phosphatidylethanol (PEth) levels reﬂecting different drinking habits in comparison to the alcohol use disorders identiﬁcation test - C (AUDIT-C). Drug Alcohol Depend. 2017;178:80–6.https://doi. org/10.1016/j.drugalcdep.2017.04.026.

[5] Isaksson A, Walther L, Hansson T, Andersson A, Alling C. Phosphatidylethanol in blood (B-PEth): a marker for alcohol use and abuse. Drug Test Anal. 2011;3(4): 195–200.https://doi.org/10.1002/dta.278.

[6] Wurst FM, Thon N, Yegles M, Schruck A, Preuss UW, Weinmann W. Ethanol metab-olites: their role in the assessment of alcohol intake. Alcohol Clin Exp Res. 2015;39 (11):2060–72.https://doi.org/10.1111/acer.12851.

[7] Wurst FM, Thon N, Aradottir S, Hartmann S, Wiesbeck GA, Lesch O, et al. Phosphatidylethanol: normalization during detoxiﬁcation, gender aspects and cor-relation with other biomarkers and self-reports. Addict Biol. 2010;15(1):88–95.

https://doi.org/10.1111/j.1369-1600.2009.00185.x.

[8] Schrieks IC, Heil AL, Hendriks HF, Mukamal KJ, Beulens JW. The effect of alcohol con-sumption on insulin sensitivity and glycemic status: a systematic review and meta-analysis of intervention studies. Diabetes Care. 2015;38(4):723–32.https://doi.org/ 10.2337/dc14-1556.

[9] Kechagias S, Nasr P, Blomdahl J, Ekstedt M. Established and emerging factors affect-ing the progression of nonalcoholic fatty liver disease. Metabolism. 2020;154183.

https://doi.org/10.1016/j.metabol.2020.154183.

[10] Sookoian S, Castano GO, Pirola CJ. Modest alcohol consumption decreases the risk of non-alcoholic fatty liver disease: a meta-analysis of 43 175 individuals. Gut. 2014;63 (3):530–2.https://doi.org/10.1136/gutjnl-2013-305718.

[11] Cao G, Yi T, Liu Q, Wang M, Tang S. Alcohol consumption and risk of fatty liver dis-ease: a meta-analysis. PeerJ. 2016;4:e2633.https://doi.org/10.7717/peerj.2633. [12] Moriya A, Iwasaki Y, Ohguchi S, Kayashima E, Mitsumune T, Taniguchi H, et al. Roles

of alcohol consumption in fatty liver: a longitudinal study. J Hepatol. 2015;62(4): 921–7.https://doi.org/10.1016/j.jhep.2014.11.025.

[13] Vilar-Gomez E, Sookoian S, Pirola CJ, Liang T, Gawrieh S, Cummings O, et al. ADH1B *2 is associated with reduced severity of nonalcoholic fatty liver disease in adults,

(5)

independent of alcohol consumption. Gastroenterology. 2020;159(3):929–43.

https://doi.org/10.1053/j.gastro.2020.05.054.

[14] Sogabe M, Okahisa T, Taniguchi T, Tomonari T, Tanaka T, Tanaka H, et al. Light alco-hol consumption plays a protective role against non-alcoalco-holic fatty liver disease in Japanese men with metabolic syndrome. Liver Int. 2015;35(6):1707–14.https:// doi.org/10.1111/liv.12754.

[15] Hamaguchi M, Kojima T, Ohbora A, Takeda N, Fukui M, Kato T. Protective effect of al-cohol consumption for fatty liver but not metabolic syndrome. World J Gastroenterol. 2012;18(2):156–67.https://doi.org/10.3748/wjg.v18.i2.156. [16] Kachele M, Wolff S, Kratzer W, Haenle M, Homann J, Trischler G, et al. Presence of fatty

liver and the relationship between alcohol consumption and markers of inﬂammation. Mediators Inﬂamm. 2015;2015:278785.https://doi.org/10.1155/2015/278785. [17] Hashimoto Y, Hamaguchi M, Kojima T, Ohshima Y, Ohbora A, Kato T, et al. Modest

alcohol consumption reduces the incidence of fatty liver in men: a population-based large-scale cohort study. J Gastroenterol Hepatol. 2015;30(3):546–52.

https://doi.org/10.1111/jgh.12786.

[18] Gunji T, Matsuhashi N, Sato H, Fujibayashi K, Okumura M, Sasabe N, et al. Light and moderate alcohol consumption signiﬁcantly reduces the prevalence of fatty liver in the Japanese male population. Am J Gastroenterol. 2009;104(9):2189–95.https:// doi.org/10.1038/ajg.2009.361.

[19] Dulai PS, Singh S, Patel J, Soni M, Prokop LJ, Younossi Z, et al. Increased risk of mor-tality byﬁbrosis stage in nonalcoholic fatty liver disease: systematic review and meta-analysis. Hepatology (Baltimore, Md). 2017;65(5):1557–65. doi:https://doi. org/10.1002/hep.29085.

[20] Perakakis N, Polyzos SA, Yazdani A, Sala-Vila A, Kountouras J, Anastasilakis AD, et al. Non-invasive diagnosis of non-alcoholic steatohepatitis andﬁbrosis with the use of omics and supervised learning: a proof of concept study. Metabolism. 2019;101: 154005.https://doi.org/10.1016/j.metabol.2019.154005.

[21] Miele L, Zocco MA, Pizzolante F, De Matthaeis N, Ainora ME, Liguori A, et al. Use of imaging techniques for non-invasive assessment in the diagnosis and staging of non-alcoholic fatty liver disease. Metabolism. 2020;154355.https://doi.org/10. 1016/j.metabol.2020.154355.