Genetic variants in the MTHFR are not associated with fatty liver disease.

Kokoteksti

(1)UEF//eRepository DSpace Rinnakkaistallenteet. https://erepo.uef.fi Terveystieteiden tiedekunta. 2020. Genetic variants in the MTHFR are not associated with fatty liver disease. De Vincentis, Antonio Wiley Tieteelliset aikakauslehtiartikkelit © 2020 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd All rights reserved http://dx.doi.org/10.1111/liv.14543 https://erepo.uef.fi/handle/123456789/8297 Downloaded from University of Eastern Finland's eRepository.

(2) Accepted Article. DR. ANTONIO DE VINCENTIS (Orcid ID : 0000-0003-0220-0500) DR. VILLE T. MÄNNISTÖ (Orcid ID : 0000-0002-0735-400X) DR. SALVATORE PETTA (Orcid ID : 0000-0002-0822-9673) DR. PAOLA DONGIOVANNI (Orcid ID : 0000-0003-4343-7213) DR. ANNA LUDOVICA FRACANZANI (Orcid ID : 0000-0001-5918-0171) DR. LUCA VALENTI (Orcid ID : 0000-0001-8909-0345) DR. STEFANO ROMEO (Orcid ID : 0000-0001-9168-4898) DR. UMBERTO VESPASIANI GENTILUCCI (Orcid ID : 0000-0002-1138-1967). Article type. : Brief Definitive Report. Handling editor: Raúl Andrade. Brief Definitive Report. Genetic variants in the MTHFR are not associated with fatty liver disease. Antonio De Vincentis*1, Rosellina Margherita Mancina*2, Jussi Philamajäki3-4, Ville Männistö 5-6, Salvatore Petta7, Paola Dongiovanni8, Anna Fracanzani8-9, Luca Valenti9-10, Federica Tavaglione12,. Stefano Romeo2,11-12 and Umberto Vespasiani-Gentilucci1. * These authors equally contributed to the present study and share first authorship. 1Department. of Internal Medicine and Geriatrics, University Campus Bio-Medico of Rome, Italy;. 2Department. of Molecular and Clinical Medicine, University of Gothenburg, Sweden; 3Clinical. Nutrition and Obesity Center, Kuopio University Hospital, Kuopio, Finland; 4Department of Public Health and Clinical Nutrition, University of Eastern Finland, Kuopio, Finland; 5Department of Medicine, University of Eastern Finland, Kuopio, Finland; 6Department of Medicine, Kuopio University Hospital, Kuopio, Finland; 7Department of Gastroenterology, Università di Palermo, This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/liv.14543 This article is protected by copyright. All rights reserved.

(3) Accepted Article. Palermo, Italy; 8General Medicine and Metabolic Diseases, Fondazione IRCCS Ca’ Granda Ospedale Policlinico Milano, Milan, Italy; 9Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy;. 10Translational. Medicine – Department of. Transfusion Medicine and Hematology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy; 11Clinical Nutrition Unit, Department of Medical and Surgical Sciences, Magna Graecia University, Catanzaro, Italy;. 12Cardiology. Department, Sahlgrenska University. Hospital, Gothenburg, Sweden Corresponding author Prof. Umberto Vespasiani Gentilucci Department of Internal Medicine and Geriatrics, University Campus Bio-Medico of Rome, via Alvaro del Portillo, 200, 00128 Rome, Italy +39 06 22541 1207; u.vespasiani@unicampus.it Word count (including title page, abstract, main text and references): 2740 Number of Figures: 0 Number of Tables: 2 Funding statement: Nothing Conflict of interest statement: Nothing to disclose ORCID. Antonio De Vincentis: 0000-0003-0220-0500; Rosellina Margherita Mancina: 0000-0002-11263071; Ville Männistö: 0000-0002-0735-400X; Salvatore Petta: 0000-0002-0822-9673; Paola Dongiovanni: 0000-0003-4343-7213; Anna Fracanzani: 0000-0001-5918-0171; Luca Valenti: 0000-0001-8909-0345; Stefano Romeo: 0000-0001-9168-4898; Umberto Vespasiani-Gentilucci: 0000-0002-1138-1967. This article is protected by copyright. All rights reserved.

(4) Accepted Article. ABSTRACT (max 150w) The common missense sequence variants of methylenetetrahydrofolate-reductase (MTHFR), rs1801131. (c.A1298C). and. rs1801133. (c.C677T),. favor. the. development. of. hyperhomocysteinemia and diminished DNA methylation. Previous studies, carried out in small series and with suboptimal characterization of the hepatic phenotype, tested the association of these genetic variants with fatty liver disease (FLD), with conflicting results. Here, we assessed the association of rs1801131 and rs1801133 with hepatic phenotype in the Liver Biopsy CrossSectional Cohort, a large cohort (n=1375 from Italy and 411 from Finland) of European individuals with suspect FLD associated with dysmetabolism. A total of 1786 subjects were analyzed by ordinal regression analyses. The rs1801131 and the rs1801133 variants were not associated with steatosis, inflammation, ballooning, or fibrosis. The present study suggests that changes in folate and methionine metabolism resulting from these 2 variants are not associated with a clinically significant impact on FLD in Europeans.. This article is protected by copyright. All rights reserved.

(5) Accepted Article. INTRODUCTION AND AIMS Several lines of evidence suggest that alterations in folate and methionine metabolism are associated with fatty liver disease (FLD) [1]. Mice fed methionine/choline deficient diet develop non-alcoholic steatohepatitis (NASH) and fibrosis [2]. In humans, hyperhomocysteinemia causes oxidative stress, inflammation, and steatosis [3-4], and drugs interfering with folate metabolism, e.g. methotrexate, induce fatty liver degeneration [5]. Methylenetetrahydrofolate-reductase (MTHFR) catalyzes the conversion of 5,10-. methylentetrahydrofolate into 5-methyltetrahydrofolate, which is essential for remethylation of homocysteine to methionine. MTHFR-knockout mice have severe liver steatosis [6], that is also seen in individuals with severe MTHFR deficiency [7]. The two most common missense genetic variants of MTHFR, rs1801131 (c.A1298C) and. rs1801133 (c. C677T), reduce enzyme activity under certain conditions and increase the susceptibility towards the development of hyperhomocysteinemia. In a recent work, Sun MY et al. reported positive associations between these 2 variants and FLD after metanalysis of several studies with conflicting results [8]. However, study samples were limited, and FLD assessment was mostly suboptimal (imaging rather than histology). Here, we aimed to assess the association of the MTHFR rs1801131 and rs1801133 variants with FLD associated with dysmetabolism in a large cohort of European individuals with available liver biopsy.. METHODS The Liver Biopsy Cross-Sectional Cohort has been previously described [9-10]. Briefly, individuals were consecutively enrolled in four European centers: the Metabolic Liver Diseases outpatient service at the Fondazione IRCCS Ca’ Granda Ospedale Policlinico Milano, Milan, Italy; the Gastrointestinal & Liver Unit of the Palermo University Hospital, Palermo, Italy; the Northern Savo Hospital District, Kuopio, Finland; the Hepatology Unit of the Campus BioMedico University Hospital, Rome, Italy. The inclusion criteria were liver biopsy for suspected non-alcoholic steatohepatitis (NASH) or severe obesity (in this case, at the time of bariatric surgery), availability of DNA samples and clinical data. Individuals with increased alcohol intake (men, >30 g/day; women, >20 g/day), viral, autoimmune hepatitis or other causes of liver disease were excluded.. This article is protected by copyright. All rights reserved.

(6) Accepted Article. The MTHFR rs1801131 and rs1801133 genotyping was performed in duplicate by TaqMan 5’-nuclease assays (Life Technologies, Carlsbad, CA). Minor allele frequency (MAF) in the overall Liver Biopsy Cross-Sectional Cohort was 0.31 and 0.39 for rs1801131 and rs1801133, respectively. For both variants, Hardy-Weinberg equilibrium was not preserved (P<0.0001). For descriptive purposes, continuous variables were shown as mean and standard deviation. or median and interquartile range. Categorical variables were presented as number and proportion. All genetic analyses were performed assuming an additive model (by coding the genotypes as 0, 1, and 2 for wild-type homozygotes, heterozygotes, and alternate allele homozygotes, respectively). Subjects were further stratified in 4 categories based on histological NAFLD activity score. (NAS) [11] as follows: 1) no steatosis; 2) steatosis +/- nonspecific inflammation (steatosis 1-3 with either inflammation 1-3 or ballooning 1-2 or none, excluding F3-F4 fibrosis); 3) NASH (steatosis 1-3 with inflammation 1-3 and ballooning 1-2, excluding F3-F4 fibrosis); 4) advanced fibrosis (F3-F4 fibrosis). The associations between the MTHFR rs1801131 or the rs1801133 variants and disease. category, the components of the NAS (severity of steatosis, necro-inflammation, and ballooning) [11], and liver fibrosis, were tested by logistic or ordinal regression analysis after adjusting for potential confounders: age, sex, body mass index (BMI), nationality (Italian Vs Finnish), indication for liver biopsy (suspected NASH Vs severe obesity), presence of impaired fasting glucose (IFG)/type-2 diabetes mellitus (T2DM), number of patatin-like phospholipase domain-. containing protein 3 (PNPLA3) I148M alleles, and presence of the transmembrane 6 superfamily member 2 (TM6SF2) E167K variant. Since not normally distributed, BMI was log-transformed before entering multivariable models. Finally, the combined effect of the two MTHFR variants, determined by the total number of risk alleles (from 0 to 4), on the histological traits of FLD was also tested by ordinal regression analysis adjusted for all the above-mentioned confounders. All analyses were performed using R statistics version 3.6.1. RESULTS As presented in Table 1), the study population included a total of 1786 individuals, of. which 1375 were Italian and 411 were Finnish. Approximately the same proportion of subjects underwent liver biopsy due to suspected NASH or due to severe obesity. Among these a total of 16% liver biopsies had no steatosis, approximately 50% had steatosis with-/-out nonspecific inflammation, 25% had the histological criteria for NASH, and 10% had advanced fibrosis.. This article is protected by copyright. All rights reserved.

(7) Accepted Article. The prevalence of the minor allele was higher in Italians than in Finns for rs1801133 (0.44 Vs 0.24, respectively, p<0.001), while a borderline difference in the allele distribution was observed for rs1801131 (0.31 Vs 0.34, respectively, p=0.08). Therefore, we decided to examine. the genetic associations stratified by European country of origin (Table 2). Firstly, allele frequencies and genotype distributions of our cohort of Italian and Finnish. descent were compared with those reported in 1000 Genomes (1000G) Project [12] for “Toscani in Italia” (TSI, 107 subjects; rs1801131 and rs180113 minor allele frequency: 0.313 and 0.467, respectively;. genotype. distribution:. AA/AC/CC. 48.6%/40.2%/11.2%. and. CC/CT/TT. 30.8%/44.9%/24.3%, respectively), and “Finnish in Finland” (FIN, 99 subjects; rs1801131 and rs180113 minor allele frequency: 0.318 and 0.273, respectively; genotype distribution: AA/AC/CC 45.5%/45.5%/9.1% and CC/CT/TT 51.5%/42.4%/6.1%, respectively), respectively, and no significant difference was observed (p value from 2 test > 0.05 for all comparisons, for. both MTHFR rs1801131 and rs1801133). After adjusting for age, sex, BMI, impaired fasting glucose (IFG)/type-2 diabetes mellitus. (T2DM), number of PNPLA3 I148M alleles, and presence of the TM6SF2 E167K variant, there. was no difference in anthropometric and clinical features of Italian and Finnish individuals when stratified by genotypes for both variants, exception made for a reduction in triglycerides in Finnish carriers of the rs1801131 variant (Table 2). Moreover, there was no difference in the distribution of metabolic diseases and in the distribution (any histological grade versus absence) of liver steatosis, inflammation, ballooning or fibrosis among genotypes (Table 2). By logistic regression analysis adjusted for confounders, including nationality and. indication to liver biopsy, neither the rs1801131 nor the rs1801133 variant were associated with FLD when compared with subjects with no steatosis [1.12 (0.90-1.40), and 1.01 (0.80-1.27), respectively]. Consistently, by ordinal regression analyses adjusted for these same confounders, neither the rs1801131 nor the rs1801133 variant were associated with steatosis [aOR 0.94 (0.821.06), and aOR 1.07 (0.94-1.21), respectively], inflammation [aOR 0.9 (0.79-1.03), and aOR 0.94 (0.83-1.08), respectively], ballooning [aOR 0.88 (0.75-1.02), and aOR 1.11 (0.96-1.29), respectively], fibrosis [aOR 1.0 (0.88-1.14), and aOR 0.92 (0.81-1.05), respectively], and disease category [aOR 0.99 (0.87-1.14), and 1.08 (0.94-1.23), respectively]. Finally, not even the total number of risk alleles (from 0 to 4) was associated with any. histological trait of FLD [aOR 1.0 (0.88-1.14) for steatosis; aOR 0.84 (0.70-1.01) for. This article is protected by copyright. All rights reserved.

(8) Accepted Article. inflammation; aOR 0.98 (0.83-1.15) for ballooning; aOR 0.91 (0.8-1.05) for fibrosis], or with liver disease category [aOR 1.08 (0.94-1.24)] DISCUSSION Disturbances in folate and methionine metabolism are a risk factor for FLD [1-5]. Genetic variations in the MTHFR (rs1801131 and rs1801133) result in changes in folate and methionine levels but conflicting results emerged from previous studies examining the effect these genetic variants on FLD [8]. The reason for these discrepancies is probably due to the relatively small sample size of these cohorts and to the use of proxies for establishing the presence of FLD rather than liver biopsy. In this study, we aimed to test the association between the MTHFR genetic variants and FLD in a large cross sectional cohort of individuals (N=1786) for whom liver biopsy was available, the Liver Biopsy Cohort. Our results robustly demonstrate that the two more common missense variants that confer a reduction of MTHFR activity, i.e., the rs1801131 and rs1801133, are not associated with the histological traits of non-alcoholic FLD. Common human genetic variants in several genes are associated with FLD [13].. Heritability for hepatic fat content is estimated to range from 20% to 70% and genetic susceptibility is therefore emerging as a possible tool for risk stratification in the current nonalcoholic FLD epidemics [13-14]. MTHFR plays a central role in the methylation of homocysteine to methionine. The rs1801133 is a c.C677T nucleotide change resulting in an alanine-to-valine substitution in protein at position 222. This variant reduces MTHFR activity in homozygotes (60%) and heterozygotes (30%), resulting in elevation of plasma homocysteine levels and diminished DNA methylation [15]. The rs1801131 variant is a c.A1298C resulting in a glutamateto-alanine change in position 429 of the protein [16]. This variant is also associated with diminished enzymatic activity and increased homocysteine levels, but to a lesser extent compared to the rs1801133 [16]. Homocysteine is readily oxidized in plasma, and during oxidation of the sulfhydryl group,. superoxide anion radical and hydrogen peroxide are generated, possibly leading to lipid peroxidation [17]. Moreover, DNA methylation, an essential epigenetic feature of DNA that modulates gene expression and genomic integrity during cellular differentiation, utilizes the universal methyl donor S-adenosyl-methionine, which depends on methionine levels. Severe impairment of MTHFR activity is well-known to promote steatosis in experimental models [6] and in humans [7]. We did not observe any association between rs1801131 or rs1801133 and FLD. An. This article is protected by copyright. All rights reserved.

(9) Accepted Article. explanation for the lack of association may be that the level of enzyme deficiency determined by these two variants is not sufficient to promote the FLD due to the redundancy of pathways metabolizing homocysteine and regenerating methionine [1]. A limitation of this study is the absence of measurement of serum homocysteine levels.. However, the rationale linking MTHFR deficiency with FLD is based not only on hyperhomocysteinemia but also on DNA methylation changes. Another limitation of the study is the absence of an study group of individuals with no liver disease. However, allele frequencies and genotype distributions of both MTHFR variants were not different with those reported in. general populations of same ancestry in the 1000G project. Moreover, results were similar when comparing the subgroup of subjects without any steatosis with those presenting FLD. Notably, this study was underpowered (power <80%) to detect small associations, in the range of those with an OR below 1.2. However, being well-powered to detect ORs 1.4, such as that reported in the meta-analysis by Sun et al. [8]), this study should be considered as quite conclusive in excluding clinically-significant associations. CONCLUSIONS The present study sheds light on a debated topic, showing no associations between the MTHFR rs1801131 or the rs1801133 variants and FLD histological traits in individuals of European descent. The present study suggests that changes in folate and methionine metabolism resulting from these 2 variants are not associated with a clinically significant impact on FLD in Europeans.. This article is protected by copyright. All rights reserved.

(10) Accepted Article. AKNOWLEDGEMENTS We thank the pathologist Vesa Kärjä for the histological evaluation of the Finnish cohort, and the gastrointestinal surgeons Sari Venesmaa and Pirjo Käkelä for collecting liver samples of the Finnish cohort.. This article is protected by copyright. All rights reserved.

(11) Accepted Article. REFERENCES 1) Mato JM, Martínez-Chantar ML, Lu SC. Methionine metabolism and liver disease. Annu Rev Nutr 2008;28:273-93.. 2) Santhekadur PK, Kumar DP, Sanyal AJ. Preclinical models of non-alcoholic fatty liver disease. J Hepatol 2018;68:230-237.. 3) Austin RC, Lentz SR, Werstuck GH. Role of hyperhomocysteinemia in endothelial dysfunction and atherothrombotic disease. Cell Death Differ 2004; S56-64.. 4) Gulsen M, Yesilova Z, Bagci S, Uygun A, Ozcan A, Ercin CN, Erdil A, Sanisoglu SY, Cakir E, Ates Y, Erbil MK, Karaeren N, Dagalp K. Elevated plasma homocysteine concentrations as a predictor of steatohepatitis in patients with non-alcoholic fatty liver disease. J Gastroenterol Hepatol 2005;20:1448-55.. 5) Patel V, Sanyal AJ. Drug-induced steatohepatitis. Clin Liver Dis 2013;17:533-46. 6) Chen Z, Karaplis AC, Ackerman SL, Pogribny IP, Melnyk S, Lussier-Cacan S, Chen MF, Pai A, John SW, Smith RS, Bottiglieri T, Bagley P, Selhub J, Rudnicki MA, James SJ, Rozen R. Mice deficient in methylenetetrahydrofolate reductase exhibit hyperhomocysteinemia and decreased methylation capacity, with neuropathology and aortic lipid deposition. Hum Mol Genet 20011;10:433-43.. 7) Mudd SH, Finklestein JD, Irreverre F, Laster L. Homocystinuria: an anzymatic defect. Science 1964;143(3613):1443-5.. 8) Sun MY, Zhang L, Shi SL, Lin JN. Associations between methylenetetrahydrofolate reductase (MTHFR) polymorphisms and non-alcoholic fatty liver disease (NAFLD) risk: a metaAnalysis. PLoS One 2016;11:e0154337.. 9) Dongiovanni P, Petta S, Maglio C, Fracanzani AL, Pipitone R, Mozzi E, Motta BM, Kaminska D, Rametta R, Grimaudo S, Pelusi S, Montalcini T, Alisi A, Maggioni M, Kärjä V, Borén J, Käkelä P, Di Marco V, Xing C, Nobili V, Dallapiccola B, Craxi A, Pihlajamäki J, Fargion S, Sjöström L, Carlsson LM, Romeo S, Valenti L. Transmembrane 6 superfamily member 2 gene variant disentangles nonalcoholic steatohepatitis from cardiovascular disease. Hepatology 2015;61:506-14.. 10) Mancina RM, Dongiovanni P, Petta S, Pingitore P, Meroni M, Rametta R, Borén J, Montalcini T, Pujia A, Wiklund O, Hindy G, Spagnuolo R, Motta BM, Pipitone RM, Craxì A, Fargion S, Nobili V, Käkelä P, Kärjä V, Männistö V, Pihlajamäki J, Reilly DF, Castro-Perez J, Kozlitina J, Valenti L, Romeo S. The MBOAT7-TMC4 Variant rs641738 Increases Risk of. This article is protected by copyright. All rights reserved.

(12) Accepted Article. Nonalcoholic Fatty Liver Disease in Individuals of European Descent. Gastroenterology 2016;150:1219-1230.. 11) Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, Ferrell LD, Liu YC, Torbenson MS, Unalp-Arida A, Yeh M, McCullough AJ, Sanyal AJ; Nonalcoholic Steatohepatitis Clinical Research Network. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005 ;41:1313-21.. 12) Auton A, Brooks LD, Durbin RM, et al. A global reference for human genetic variation. Nature. 2015;526(7571):68-74.. 13) Romeo S, Sanyal A, Valenti L. Leveraging human genetics to identify potential new treatments for fatty liver disease. Cell Metab 2020;31:35-45.. 14) Vespasiani-Gentilucci U, Gallo P, Dell'Unto C, Volpentesta M, Antonelli-Incalzi R, Picardi A. Promoting genetics in non-alcoholic fatty liver disease: Combined risk score through polymorphisms and clinical variables. World J Gastroenterol 2018;24:4835-4845.. 15) Frosst P, Blom HJ, Milos R, Goyette P, Sheppard CA, Matthews RG, Boers GJ, den Heijer M, Kluijtmans LA, van den Heuvel LP, et al. A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat Genet 1995;10:111-3.. 16) Weisberg I, Tran P, Christensen B, Sibani S, Rozen R. A second genetic polymorphism in methylenetetrahydrofolate reductase (MTHFR) associated with decreased enzyme activity. Mol Genet Metab. 1998 Jul;64(3):169-72. PubMed PMID: 9719624.. 17) Hayden MR, Tyagi SC. Homocysteine and reactive oxygen species in metabolic syndrome, type 2 diabetes mellitus, and atheroscleropathy: the pleiotropic effects of folate supplementation. Nutr J. 2004 May 10;3:4.. This article is protected by copyright. All rights reserved.

(13) Accepted Article. Table 1. General characteristics of the study population N. 1786. Age (years). 47.3 (11.8). Men, n (%). 929 (52%). BMI (Kg/m2). 35.2 [28.3-41.8]. Diabetes Mellitus, n (%). 442 (25%). Liver Biopsy for NASH, n (%). 898 (50.3%). Ethnicity, n(%) Italian. 1375 (77%). Finnish. 411 (23%). Glycaemia (mg/dL). 97 [88-111]. Insulin (µU/L). 15.6 [10.7-22.4]. HOMA-IR. 3.8 [2.4-5.7]. Total cholesterol (mg/dL). 192.2 (43.3). LDL (mg/dL). 114.4 (37.9). HDL (mg/dL). 48.4 (14.5). Triglycerides (mg/dL). 123 [88-171]. ALT (UI/L). 39 [23-65]. MTHFR 133, n(%) AA. 694 (39%). AC. 781 (44%). CC. 311 (17%). MTHFR 133 C allele frequency. 0.393. MTHFR 131, n(%). MTHFR 131 T allele frequency. CC. 887 (50%). CT. 677 (38%). TT. 222 (12%) 0.314. Steatosis (presence), n (%). 1504 (84.2%). Inflammation (presence), n (%). 1098 (61.5%). This article is protected by copyright. All rights reserved.

(14) 653 (36.6%). Fibrosis (presence), n (%). 1014 (56.8%). Accepted Article. Ballooning (presence), n (%) Liver disease category, n (%) No steatosis (controls). 282 (16%). Steatosis with-/-out non-specific inflammation. 884 (49%). NASH. 429 (24%). Advanced fibrosis. 191 (11%). Data were shown as mean (standard deviation) or median [interquartile range] for continuous traits, if normally or non-normally distributed, respectively; categorical traits were shown as number and proportion (%).. This article is protected by copyright. All rights reserved.

(15) Accepted Article. Table 2. Demographic, anthropometric, and clinical features of the Liver Biopsy Cohort stratified by European Country of origin and by MTHFR rs1801131 and rs1801133 genotypes. MTHFR rs1801131. Italian descent. MTHFR rs1801133. AA. AC. CC. N. 700 (50.9%). 506 (36.8%). 169 (12.3%). Age (years). 46.4 (12.7). 47.7 (12.1). 47.7 (12.7). Men, n (%). 323 (46.1%). 228 (45.1%). 31.6 [27.5-. 31.2 [27.1-. 38.9]. 38.7]. Diabetes Mellitus, n (%). 146 (21.1%). Liver Biopsy for NASH, n (%). 453 (64.7%). BMI. (Kg/m2). Glycaemia (mg/dL) Insulin (µU/L). 93 [85-106]. CC. CT. TT. p. 453 (32.9%). 635 (46.2%). 287 (20.9%). -. 0.205. 47.2 (12.3). 47.4 (12.3). 46 (13). 0.254. 83 (49.1%). 0.67. 230 (50.8%). 272 (42.9%). 132 (46%). 0.037. 31.4 [27-40]. 0.831. 32.6 [27.5-. 31.1 [27.2-. 31.3 [27-. 39.6]. 38.4]. 38.7]. 122 (24.5%). 38 (22.8%). 0.39. 99 (22.2%). 149 (23.7%). 58 (20.5%). 0.548. 340 (67.2%). 105 (62.1%). 0.414. 284 (62.7%). 424 (66.8%). 190 (66.2%). 0.503. 94 [87-109.1]. 0.627. 94 [87-109]. 95 [86-108]. 16 [9.6-21]. 0.945. 16 [10.8-. 14.9 [10.4-. 16.5 [10.6-. 22.1]. 21]. 23]. 95 [86.7106.9]. p. 91 [84.1101.8]. 0.627. 0.185. 15.1 [10.7-. 15 [10.7-. 22.9]. 21.6]. HOMA-IR. 3.6 [2.3-5.7]. 3.5 [2.4-5.2]. 3.4 [2.3-4.9]. 0.601. 3.6 [2.4-5.2]. 3.4 [2.3-5.4]. 3.7 [2.3-5.4]. 0.940. Total cholesterol (mg/dL). 201.5 (41.6). 201.4 (42.1). 201.3 (38.1). 0.973. 201.1 (41.9). 201.3 (41). 202.3 (41.6). 0.478. LDL (mg/dL). 121.7 (36.2). 122.5 (37.7). 120.9 (36.1). 0.818. 120.3 (36.4). 122.6 (37.3). 123.1 (36.2). 0.275. HDL (mg/dL). 49.7 (15.3). 49.7 (14.8). 50.8 (14.8). 0.754. 51.7 (16.3). 49 (13.5). 48.6 (16). 0.06. This article is protected by copyright. All rights reserved. 0.731.

(16) Accepted Article. Triglycerides (mg/dL). 127.5 [89.0-. 121.3 [87.3-. 122.0 [86.0-. 118.0 [83.3-. 125.0 [90.1-. 132 [92.1-. 174.0]. 178.0]. 168.3]. 166.8]. 178.0]. 179.8]. 42 [22-72]. 43 [22-69[. 0.140. 39 [22-65]. 42 [22-70]. 43 [24-76.5]. 0.167. Steatosis (presence), n (%). 639 (91.3%). 473 (93.5%). 159 (94.1%). 0.118. 415 (91.6%). 592 (93.2%). 264 (92%). 0.977. Inflammation (presence), n (%). 494 (70.6%). 366 (72.3%). 107 (63.3%). 0.096. 314 (69.3%). 449 (70.7%). 204 (71.1%). 0.681. Ballooning (presence), n (%). 289 (41.3%). 207 (40.9%). 60 (35.5%). 0.374. 172 (38%). 255 (40.2%). 129 (44.9%). 0.166. Fibrosis (presence), n (%). 421 (60.1%). 311 (61.5%). 106 (62.7%). 0.867. 283 (62.5%). 376 (59.2%). 179 (62.4%). 0.852. No steatosis. 61 (8.7%). 33 (6.5%). 10 (5.9%). 38 (8.4%). 43 (6.8%). 23 (8%). Steatosis w/o ns infl. 364 (52%). 273 (54%). 97 (57.4%). 247 (54.5%). 350 (55.1%). 137 (47.7%). NASH. 192 (27.4%). 124 (24.5%). 40 (23.7%). 114 (25.2%). 151 (23.8%). 91 (31.7%). Advanced fibrosis. 83 (11.9%). 76 (15%). 22 (13%). 54 (11.9%). 91 (14.3%). 36 (12.5%). AA. AC. CC. p. CC. CT. TT. p. 187 (45.5%). 171 (41.6%). 53 (12.9%). -. 241 (58.6%). 146 (35.5%). 24 (5.9%). -. Age (years). 48.7 (9). 47.9 (9). 47.9 (9.3). 0.566. 48.2 (9). 48.3 (9.1). 48.8 (9.4). 0.92. Men, n (%). 131 (70.1%). 118 (69%). 46 (86.8%). 0.033. 174 (72.2%). 108 (74%). 13 (54.2%). 0.132. 42 [38.5-. 42.5 [39-. 41.9 [38.7-. 42.1 [39-. 42.5 [38.4-. 44.8 [39.6-. 46.8]. 46.6]. 44.4]. 45.8]. 46.8]. 47.3]. 60 (32.1%). 56 (32.7%). 20 (37.7%). 0.737. 77 (32%). 49 (33.6%). 10 (41.7%). 0.621. 0 (0%). 0 (0%). 0 (0%). 1. 0 (0%). 0 (0%). 0 (0%). 1. ALT (UI/L). 38 [22.860.2]. 0.538. 0.113. Liver disease category, n(%). Finnish descent N. BMI (Kg/m2) Diabetes Mellitus, n (%) Liver Biopsy for NASH, n (%). This article is protected by copyright. All rights reserved. 0.767. 0.383. 0.081. 0.337.

(17) Accepted Article. Glycaemia (mg/dL). 108 [99-. 104.4 [95.4-. 105.3 [93.6-. 106.2 [97.2-. 106.2 [95.4-. 107.1 [98.5-. 118.8]. 122.4]. 122.4]. 118.8]. 120.6]. 116.1]. 16.5 [12.2-. 16 [10.2-. 12.6 [8.6-. 15.2 [10.2-. 16.1 [11.4-. 20.1 [15.5-. 25.1]. 22.6]. 19.5]. 22.9]. 23.1]. 25.8]. HOMA-IR. 4.6 [3.1-7.3]. 4.2 [2.5-6.5]. 3.2 [2.3-5.1]. 0.392. 4.2 [2.6-6.6]. 4.4 [2.8-6.6]. 5 [3.9-7.5]. 0.788. Total cholesterol (mg/dL). 166.9 (36.8). 161.2 (35.1). 159.8 (38.1). 0.199. 164.1 (36.8). 162.3 (34.7). 167.1 (42.1). 0.980. LDL (mg/dL). 94.2 (32.9). 91.1 (30.7). 90.2 (35.2). 0.512. 92.7 (32.4). 91.4 (31). 95.3 (39.9). 0.824. HDL (mg/dL). 44.1 (11.6). 43.8 (11.9). 46.1 (11.3). 0.675. 43.9 (11). 44.8 (12.2). 43.7 (15.6). 0.345. Triglycerides (mg/dL). 123.6 [94.8-. 117.8 [88.6-. 103.6 [81.7-. 122.2 [89.5-. 116.9 [90.3-. 116.9 [93.9-. 170.5]. 157.7]. 136.8]. 161.2]. 162.1]. 167.0]. 35 [26-56]. 34 [25-46]. 31 [26-45]. 35 [25-49]. 34 [25-54.2]. 37 [29.5-57]. No steatosis. 80 (42.8%). 74 (43.3%). 24 (45.3%). 108 (44.8%). 60 (41.1%). 10 (41.7%). Steatosis with-/-out nonspecific inflammation. 72 (38.5%). 58 (33.9%). 20 (37.7%). 82 (34%). 59 (40.4%). 9 (37.5%). NASH. 30 (16%). 35 (20.5%). 8 (15.1%). 45 (18.7%). 24 (16.4%). 4 (16.7%). Advanced fibrosis. 5 (2.7%). 4 (2.3%). 1 (1.9%). 6 (2.5%). 3 (2.1%). 1 (4.2%). Steatosis (presence), n (%). 107 (57.2%). 97 (56.7%). 29 (54.7%). 0.691. 133 (55.2%). 86 (58.9%). 14 (58.3%). 0.842. Inflammation (presence), n (%). 60 (32.1%). 57 (33.3%). 14 (26.4%). 0.852. 82 (34%). 40 (27.4%). 9 (37.5%). 0.148. Ballooning (presence), n (%). 45 (24.1%). 42 (24.6%). 10 (18.9%). 0.587. 60 (24.9%). 31 (21.2%). 6 (25%). 0.248. Fibrosis (presence), n (%). 80 (42.8%). 74 (43.3%). 22 (41.5%). 0.853. 111 (46.1%). 55 (37.7%). 10 (41.7%). 0.165. Insulin (µU/L). ALT (UI/L). 0.219 0.232. 0.021 0.264. 0.159 0.705. 0.458 0.797. Liver disease category, n(%). This article is protected by copyright. All rights reserved. 0.893. 0.615.

(18) Accepted Article. Data were shown as mean (standard deviation) or median [interquartile range] for continuous traits, if normally or non-normally distributed, respectively; categorical traits were shown as number and proportion (%). P values are from linear or binary logistic regressions between each variable and different genotypes, adjusted for age, sex, logBMI, diabetes mellitus, number of PNPLA3 I148M and number of TM6SF2 E167K mutant allele. Non-normally distributed variables were log-transformed before entering the models. BMI, body mass index; HOMA-IR, homeostatic model for assessment of insulin resistance; ALT, alanine aminotransferase; HDL, high-density lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol. This article is protected by copyright. All rights reserved.

(19)