Good afternoon. I'm Anna Mae Deal from Duke University, and my co-chair is Steve Caldwell from UVA. We'd like to welcome you to Session 3, which is the Combined Impact of Alcohol and Metabolic Syndrome on Liver Disease. To get started, I'll briefly introduce our four speakers. Our first talk on alcohol consumption in NAPL will be given by Dr. Sammy Galwith. Dr. Galwith is an associate professor of medicine at the University of Indiana and an expert on clinical trials in alcohol-related liver disease in NAPL. Our second talk on dietary intervention in fatty liver disease will be given by Dr. Shiraz Zalbersaghi. Dr. Zalbersaghi is a nutritional epidemiologist and the head of the School of Public Health at the University of Haifa in Israel. Our third talk discusses synergism between alcohol and obesity on carcinogenesis. This will be given by Dr. Louis Roberts. Dr. Roberts is a professor of medicine at Mayo Clinic. He leads a research program that's defining molecular mechanisms of liver and biliary carcinogenesis. The final talk in this session is about genetic modifiers of ALD and NAPL severity. This will be given by Dr. Stefano Romeo. Dr. Romeo is a professor of molecular and clinical medicine at Saligenska Academy in Germany. He is an expert in the role of genetic variation in modulating the outcomes of metabolic liver disease. Please remember to submit your questions while the talks are ongoing, and you can do this online by using the Q&A in the chat box. Your questions will be answered subsequently in a live Q&A session. Thanks, and let's get started. Thank you. I'd also like to thank the program chairs and the ASLD for the opportunity to speak on this topic. Let us first start by discussing the threshold setting for alcohol benefit and harm in NAFLD. What actually constitutes significant alcohol consumption? Both ASLD and ease of recent guidelines rely on experts' consensus in setting these specific thresholds, and they both acknowledge the paucity of data on a specific cutoff at the time of writing the guidelines. The ASLD guidelines set the threshold at more than two drinks a day for a woman and more than three drinks a day for men, and it cites the currently outdated 1986 U.S. National Mortality Followback Survey. Now, the ESL guidelines set the cutoff at more than 20 grams a day for women and more than 30 grams a day for men, and it cites data from the Dionysus study. The NASH CRN, on the other hand, sets the threshold lower and consistent with the threshold set by the dietary guidelines for Americans at more than one drink a day for women and more than two drinks a day for men. However, the definition of alcohol content in each standard drink is different. Looking at the Dionysus study, the data in that study indicate dose-dependent increase in the risk of non-serotic liver disease and cirrhosis with increasing average alcohol consumption. At the 30 grams consumption a day, the risk of both endpoints was less than 1%, and therefore that threshold was set as the risk threshold for alcoholic liver disease in men and women. In a follow-up study, the interaction of alcohol consumption with obesity was demonstrated, and importantly in that study, the controls who had BMI less than 25 but consumed less than 30 grams of alcohol a day did have a prevalence of fat liver of 16%. Now, several studies indicate that women are at higher risk for alcohol-associated liver disease and at lower levels of consumption than men. In this study, for example, the J-shaped appearance of the risk curve for men for alcoholic cirrhosis and alcoholic liver disease is not seen in women who have higher risk for developing both endpoints at lower levels of consumption. So, this finding has been validated by many other studies summarized here in this meta-analysis where the risk of cirrhosis in women has increased in a dose-dependent fashion with any consumption of alcohol and starting to increase at consumption of one drink of alcohol a day. There are several challenges in examining the benefits and harms of moderate alcohol consumption in NAFLD. A primary challenge is the field requires the use of questionnaires. It relies on the use of questionnaires to measure the exposure to alcohol consumption, and the problem is questionnaires rely on self-reporting, which introduces recall bias. Another challenge is the inaccuracy of converting from drink numbers to grams or the use of daily or weekly averages in binge drinking. The use of recent versus lifetime drinking as a reference introduces misclassification error, as some current non-drinkers could have been past heavy drinkers, and that introduces misclassification error. Many studies use the same cutoff of alcohol consumption for men and women, and data on changing alcohol consumption habits over time, confounding covariates, and alcoholic drink ingredient or calories are frequently not captured or adjusted for in many studies. Furthermore, a standard drink alcohol content and size varies across countries. In the United States, a standard drink is defined as 14 grams of alcohol, whereas a standard drink in different countries may range from 80 grams of alcohol and up to 20 grams. Furthermore, the definition of moderate alcohol consumption varies across different studies from different countries, and it ranges from 1 to 3 drinks, or from 6 and up to 40 grams in different studies from different countries. As could be seen, the consumption level is not stratified in many studies for men and women. So with all these caveats, let us turn to examining the effect of moderate alcohol consumption on hepatic outcomes. Several observational studies evaluated the effects of MAC on NAFLD risk, and as could be seen, some have shown reduced risk with MAC. However, other studies have shown increased risk even with consumption of less than a drink a day in the setting of obesity or consumption of non-wine beverages. A recent cohort study showed increased risk of worsening hepatic fibrosis as measured by FIP4 or NAFLD fibrosis score with moderate alcohol consumption. In this study, the prevalence of fatty liver measured by MR spectroscopy was higher in modest drinkers compared to non-drinkers. Modest drinkers also had higher median liver fat content by MR spectroscopy than non-drinkers. However, there was no significant difference between significant liver stiffness measurement between the two groups. Importantly, MAC was no longer associated with fatty liver after adjustment for age, gender, and the metabolic syndrome. When one evaluates the effects of MAC on NAFLD histology, you see that the majority of studies report either neutral or beneficial effects on histology with two notable exemptions. One study used the Mendelian randomization design relying on the alcohol dehydrogenase 1B variant number 1229984, and that study demonstrated increased risk for NASH in modest drinkers who also had worse NAFLD histology. Another study using longitudinal paired biopsies demonstrated a reduced risk for NASH in modest drinkers, but modest drinkers also were less likely to improve steatosis or to resolve NASH on subsequent biopsies. We examined the effect of MAC on NAFLD histology recently, and we observed a beneficial effect of MAC on NAFLD histology. Compared to non-drinkers, modest drinkers who consumed up to 20 grams of alcohol a day had a dose-dependent reduction in the prevalence of ballooning, portal inflammation, increased fibrosis, and deafness steatohepatitis. In a multivariable model adjusting for multiple covariates, the association of MAC with lower risk of fibrosis, deafness steatohepatitis, and NASH of four or greater was confirmed. We also observed that the star 2 allele of the alpha alcohol dehydrogenase 1B variant number 1229984, this allele is associated with faster alcohol metabolism, and we also observed that this allele is associated with lower risk of ballooning, fibrosis, deafness steatohepatitis, and NASH of four or greater. When we examined the interaction of MAC with the star 2 allele, we observed that MAC offers the most significant reduction in NASH risk in those who carry the star 2 allele of the ADH1B variant compared to the star 1 allele. An important finding in many studies is that NAFLD patients who consume modest amounts of alcohol have better metabolic and socioeconomic profiles, as shown in this study. Modest drinkers have generally lower BMI, lower frequency of diabetes, better HDL profile, and they also have different diet composition. They have higher incomes and higher degree of education. This finding, echoed in many other studies that collect and report these data, raised the question if MAC indeed is a surrogate for a better socioeconomic status. The synergy of alcohol with the metabolic syndrome component increases the risk of incidence liver disease even with modest alcohol consumption. Here, modest alcohol consumption, shown in the gray, increases the risk of severe liver disease in the setting of obesity, central obesity, and with diabetes. When looked at as a continuous variable, alcohol consumption in an exponential manner and beginning at low thresholds, lower than the one set in the guidelines, increases the risk of incidence of liver disease in different BMI strata with different waist circumferences and with diabetes. Let us now examine the effects of MAC on cardiovascular outcomes. In this study from high-income countries, this systematic analysis demonstrated a significant increase in the risk of all-cause mortality with alcohol consumption of 100 grams per week. Now, the same threshold, 100 grams a week of alcohol, was associated with the maximum reduction of cardiovascular disease, as shown here at the bottom of the J-shaped curve. Now, when cardiovascular disease risk is broken down, one sees that the benefit attributable to alcohol is mainly via the reduction of the risk of myocardial infarction and coronary artery disease at the expense of considerable increase of the risk of all strokes, heart failure, and death from other cardiovascular disease, and also at lower thresholds than 100 grams per week. Looking at the effects of MAC on cardiovascular disease in NAFLD, there is limited data. In this study in Korean men who had NAFLD by ultrasound, mild and moderate drinkers had lower prevalence of carotid plaque compared to non-drinkers, and they also had lower prevalence of carotid artery stenosis compared to non-drinkers. In U.S. patients with NAFLD, these benefits were not demonstrated. Indeed, MAC did not reduce the risk of diabetes, hypertension, or dyslipidemia, which are risk factors for cardiovascular disease, nor did it reduce the risk of abnormal left ventricular relaxation or coronary artery calcification score greater than one, which are markers for subclinical cardiovascular disease. Let us now evaluate the effects of MAC on cancer. In this large systematic analysis, the life years gained by alcohol attributable reduction in coronary artery disease, shown here in purple, are offset by considerable increase in the life years lost due to alcohol attributable increase in cancers, alcoholic liver disease, alcohol-related other diseases, and injuries. This pattern is seen both in women and men and generally increases with age. Now, the relative risk for harm was calculated in that study to begin to rise with consumption of even one alcoholic beverage a day. In NAFLD, there is limited data on the risk of cancer with modest alcohol consumption. In this study of Nash cirrhotics from a single U.S. center, drinkers of any amount, including social drinking, had considerably higher risk of developing hepatocellular carcinoma compared to Nash cirrhotics who never drank alcohol. In a study of a Scottish man with long-term follow-up, the interaction of alcohol, even in modest amounts, with overweight and obesity significantly increased the risk of mortality from liver cancer or liver disease. Let us now look at the effects of MAC on mortality and other heart outcomes. In the multi-ethnic international cohort of patients with biopsy-proven NAFLD-compensated cirrhosis, moderate alcohol consumption was associated with lower transplant-free survival, higher incidence of new-onset decompensation, and higher risk and incidence of hepatocellular carcinoma compared to non-drinkers. In a recent study that evaluated data from NHANES-3, patients with fatty liver who excessively drank alcohol, shown here in the brown line, had worse survival than non-drinkers, shown in the red line. Now, non-excessive drinkers of alcohol who had fatty liver had better survival than non-drinkers in the unadjusted analysis. However, this dissipated after adjusting for the metabolic syndrome and other covariates. In the same study, patients with fatty liver who binged alcohol more than 13 times a year had considerable increase in the risk of mortality. Another study evaluated the effects of MAC on mortality in U.S. patients with NAFLD, and as shown here, compared to non-drinkers shown in red, modest drinkers who consumed half to one and a half drinks of alcohol daily had better survival, both in men and women. However, the range, the therapeutic range, was very narrow, and those who consumed more than one and a half drinks a day indeed had higher risk of mortality. This study also showed the importance of other covariates in determining the risk of death in patients with NAFLD. A recent population-based study from Finland demonstrated that baseline alcohol consumption, in fact, increases the risk of advanced liver disease in both men and women. It also increases the risk of cancer, but reduces the risk of cardiovascular disease and all-cause mortality. The reduction in all-cause mortality showed a J-shaped appearance with the maximum reduction in that risk between with consumption of zero to nine grams of alcohol daily. Importantly, in the subgroup analysis, this benefit in risk reduction was only limited to those who were never smokers, shown here in the purple line, but not informer or current smokers. So, looking at the big picture, alcohol interaction with numerous other covariates, in fact, influences NAFLD risk severity and survival, and we need to look at alcohol consumption in this context. So, in summary, emerging data show lower thresholds for hepatic and extra-hepatic harm than current guidelines cutoffs for non-significant alcohol use in NAFLD. Data showing MAC benefits on NAFLD risk and severity or heart outcomes originate from observational studies with neurotherapeutic window. The reduced risk of coronary artery disease with alcohol is offset by increased risks of cancer, advanced liver disease, and other alcohol-related diseases and injuries. Alcohol synergy with metabolic traits results in poor outcomes, and we all know that NAFLD patients are enriched with these traits. So, based on the current data, alcohol consumption, even in modest amounts, cannot be recommended to patients with NAFLD. Thank you for your attention. Thank you very much for the invitation to give this talk today on the dietary intervention in fatty liver disease, and the questions for discussion are which dietary patterns work and are evidence-based, which dietary patterns are less evidence-based but still might be effective, does macronutrient content and other compounds, like coffee and polyphenols, play a role beyond caloric deficit in the treatment of NAFLD, and finally, how to optimize success of dietary interventions. And the most evidence-based dietary pattern in the treatment of NAFLD is the Mediterranean dietary pattern, and I want to emphasize a couple of points regarding this part, this pattern. First, remember that it is actually increased fat dietary pattern, but the type of fat that we have mostly in this diet is monounsaturated fat, while the amount of saturated fat is very low in this diet. Secondly, this is actually a reduced carb diet, where only 40 percent of the calories are coming from carbohydrates. And finally, this is a dietary pattern that is based on unprocessed or minimally processed food, with very low consumption of ultra-processed food, and this is a point that will be discussed further in this talk. So, the Mediterranean dietary pattern has been demonstrated in randomized clinical trials to be effective more than other types of diet. This is a meta-analysis of studies lasting from six weeks to six months, and we can see a greater reduction in liver stiffness, fatty liver index, and HOMA insulin resistance with the Mediterranean diet compared to other diets. But in a longer follow-up, we also want to see positive effects. So, in this cohort study, with a very large sample size and follow-up of six years, we can see that people who had a diet that was similar to the Mediterranean diet, that had a greater adherence to Mediterranean diet, had, during the follow-up, actually a reduction in the amount of liver fat and lower incidence of NAFLD. In a longer-term, randomized clinical trial, we can also see beneficial effects of Mediterranean diet over low-fat diet. In this specific randomized clinical trial, they actually compared two types of Mediterranean diets to one another and to general healthy dietary guidelines. So, we can see that weight reduction during follow-up was similar between the two types of Mediterranean diet and greater than the general lifestyle guidelines treatment. What was the difference between the two types of Mediterranean diet? In fact, the main difference was that the diet called the green Mediterranean diet was also enriched with green plants and green tea, and therefore enriched with polyphenols. And what is interesting is that despite a similar weight reduction, the reduction in the amount of liver fat was greatest in the green Mediterranean diet, meaning the Mediterranean diet with further enrichment in polyphenols. And interestingly, they looked at what were the factors that predicted the best improvement in the amount of liver fat reduction. And what did we see? That people who reduced the intake of red and processed meat had the greatest reduction in the amount of liver fat, and those who had increased consumption of walnuts, which are rich in omega-3 fatty acids, had a better reduction in the amount of liver fat. And of course, those who had higher intake of polyphenols had a greater reduction in the amount of liver fat. The importance of polyphenols is also supported in this observational study we conducted in Israel. In this study, we tested the polyphenol intake of a general population, and they had an abdominal ultrasound to diagnose NAFLD and non-invasive markers to evaluate the level of fibrosis. And what we noticed is that adjusting for other risk factors, higher intake of polyphenols was related with reduced risk of NAFLD, significant fibrosis, according to markers, and reduction in insulin resistance. Just to remind you that polyphenols are found in fruits, vegetables, nuts, coffee, and tea. So that brings us to the importance of dietary composition beyond weight reduction. And this beautiful study shows us the importance of the type of fat that we eat. So, if we look in the upper figure, we can see that if we have a hyper-energetic diet, it is always good for reduction of the amount of liver fat, regardless of the type of fat. However, if we have an isoenergetic diet, and we have in this diet saturated fat, then it's not a good thing. If it's a high-fat diet, mostly rich in saturated fat, then it would increase the amount of liver fat. If we have a hyper-energetic diet, it's always bad. It will always increase the amount of liver fat, regardless of its specific macronutrient content. But as you can see in the light blue bar, if the diet is hyper-energetic but rich, not in saturated fat, but with unsaturated fat, polyunsaturated fat, then the harmful effect would be much lower than the harmful effect of saturated fat. And that brings us to look at the exact type of fat and its effect on liver fat. So, if we have a hyper-energetic diet and it is rich in saturated fat, it would increase the amount of liver fat drastically, dramatically. But if it is rich only in monounsaturated fat or polyunsaturated fat, then the increased amount of liver fat would be much lower. And also, lastly, in the upper right figure, we can see that if we have an isoenergetic or hyper-energetic diet that is similar to the Mediterranean diet, then we would have still a reduction in the amount of liver fat. In contrast, if it is rich in saturated fat, there will not be a reduction in the amount of liver fat. So, that actually supports how harmful effect of saturated fat on the liver and the beneficial effect of Mediterranean diet and diet that is rich in monounsaturated fat in terms of liver fat reduction. So, it's not surprising that the 2015 U.S. dietary guidelines actually lifted the ban on total dietary fat intake. So, we no longer discuss about the total fat intake. We only discuss about the quality of fat. And we need to optimize the type of fat that we eat, but not the total amount eaten. And that actually supports the principles of the Mediterranean diet, which, as I mentioned, is not low-fat diet. On the contrary, but the type of fat is monounsaturated fat and not saturated fat. Furthermore, if we need further support to recommend the patients on Mediterranean diet, we can look also on its association with reduced risk of hepatocellular carcinoma and chronic liver disease mortality. In this very large cohort study with 17 years follow-up, we can see that Mediterranean diet or similar type of diet, like the healthy eating index diet, are related with reduced risk of chronic liver disease mortality and reduced risk of HCC. This is also demonstrated in this figure, showing us that everything that is helpful for reduction in the amount of liver fat is also helpful for the prevention of HCC. These are all large prospective cohort studies or meta-analyses of prospective cohort studies. And so, you can see that red and processed meat are related with increased risk of HCC, fish, vegetables, fruits related with reduced risk of HCC, as well as healthy eating index diet. Coffee, I'll get back to coffee also in the next slides, but coffee is related with reduced risk of HCC. So, Mediterranean diet is also probably good for the prevention of HCC. And the polar opposite of Mediterranean diet is ultra-processed food and drinks, which are a major source of sugar and saturated fat in a Western diet. And what are actually ultra-processed foods? These are formulations of ingredients made by a series of industrial processes and are packed in a sophisticated way to be more attractive. And they are actually designed to create highly profitable products, ready-to-consume, hyper-palatable products, which are intended to replace real foods. So, how much do we eat from ultra-processed food? Apparently, about half of the calories in Western populations come from ultra-processed food. And please notice that the consumption is highest among children and adolescents in different countries, reaching, for example, in the UK, to more than 65% of the calories per day coming from ultra-processed foods. And ultra-processed food has been demonstrated to be related with chronic diseases and also with obesity, as in this meta-analysis, with obesity and abdominal obesity and overweight. And interestingly, it is also associated with abdominal obesity, as you can see on the right, in a dose-response manner. The more ultra-processed food we eat, the more we have abdominal obesity. And in this randomized clinical trial with the crossover design, people were provided with either unprocessed or minimally processed food or ultra-processed foods. And you can see examples for the ultra-processed foods, meals that they were receiving during the two-weeks intervention. And interestingly, during the time they were eating the ultra-processed food diet, they had increased caloric intake, and not surprisingly, they gained weight. This means that ultra-processed food is indeed designed to increase, to be hyper-palatable and increase consumption. And if we want our patients to have weight reduction, we need to convince them to reduce the consumption of ultra-processed foods. Ultra-processed food has also been demonstrated to be related with incidents of type 2 diabetes in the UK Biobank study. And we all know the type today because it's highly relevant for patients with NAFLD. We have shown recently a dose-response association between ultra-processed food consumption and metabolic syndrome. Every 10% increase in the consumption of ultra-processed food leads to increased risk of metabolic syndrome in general population and specifically among patients with NAFLD. More interestingly, we have found an interaction between consumption of ultra-processed food and smoking in relation to significant fibrosis. So, if people were having a high amount of ultra-processed food and they were smokers, they had about a two-fold increased risk of having significant fibrosis according to fibrosis markers. One of the explanations for the association between ultra-processed food and harmful health effects and also liver damage is their high content of advanced glycations and products, which are actually glycotoxins. They are found in high amounts in highly processed foods like processed meats and foods that are high in saturated fat and sugar, which are prepared at high temperatures for a long duration. And when we have, when we eat food that is rich with ages, we have increased serum ages, which were demonstrated to be related with oxidative stress, insulin resistance, type 2 diabetes, cardiovascular disease, NAFLD, and liver damage and even liver cancer. And maybe not surprisingly, Mediterranean type of diet is actually low in ages. So, the foods included in Mediterranean diet and home-based cooking and minimally processed foods have low ages content, and this is maybe one more reason to recommend the Mediterranean diet. Now, let's move a bit to, specifically, to sugar. Sugar is very highly, very strongly related with NAFLD. It's very evidence-based, and I would give just a few recent examples. So, children, even infants, having soft drinks on a daily basis, and soft drinks include also fruit juice, fruit concentrates, soft drinks, and lemonades. If they have a couple of those soft drinks per day, two portions per day, they have three-fold increased risk of developing NAFLD at the age of 10 years old. And in a recent study, a very large cohort study with four years follow-up, it has been demonstrated that there is a dose-response association between the number of servings of soft drinks and the incidence of NAFLD. Those who have four servings per week have almost 50% increased risk of having incidence of NAFLD. There is always a debate if fructose is worse than glucose, and this recent study shows us that fructose may be even worse than glucose, because in this study, people were provided with three times a day of soft drinks with either fructose, sucrose, and glucose. And let me just remind you, the sucrose contains 50% of glucose and 50% of fructose. So, only the fructose-containing beverages actually led to increased hepatic lipogenesis. The novel lipogenesis where the glucose-containing soft drinks led to very minor increase in hepatic lipogenesis. So, fructose does seem to have worse effect. There is always the question regarding intermittent fasting, and many patients start to do these diets that get more and more popular. So, we don't have much evidence in NAFLD, but in this small, short-term study, we can get some idea about the effect of intermittent fasting. In this study, they had either regular low-carb diet or intermittent fasting in a way that people were eating five days a week regularly, and two days a week they were almost fasting. They had five calories per day. And we can see that both diets led to similar reduction in the amount of liver fat. So, intermittent fasting was not better than the usual diet, but it can be another option for the patients. Another small, short-term study of only eight weeks in which intermittent fasting was compared to actually no intervention. In this intermittent fasting, they had a fast day and feast day. A fast day where they had very small amount of calories, and a feast day where they could eat as much as they want. And generally, this diet led to reduction in ALT levels and a reduction in the levels of shear wave elastography. But again, we need longer term and larger studies. With regard to weight reduction, it has been demonstrated that intermittent fasting as compared to regular low-calorie diet has similar effect in terms of weight reduction and in terms of reduction of risk factors for cardiovascular disease. So, it's an option, but it's not necessarily better than a regular diet for reduction of liver fat or for weight reduction. Low-carb diet, it's a very attractive option, too. So, first of all, there are a couple of levels of low-carb diet. It can be like Mediterranean diet, reduced-carb diet. It can be low-carb diet with only about 25% of the calories from carbohydrates, and it can be very low-carb diet, like a ketogenic diet. Generally, low-carb diet did not show a better effect of the reduction of liver fat than usual low-fat or other types of diet. And although it seems in a very short-term follow-up that low-carb diet is more beneficial than low-fat diet, when we keep following the patients, we see that there is no actually difference between the two types of diet in terms of liver fat reduction. Generally, the studies looking at the effect of low-carb diet are very small and with very short duration, and we cannot reach any conclusive recommendations with regard to the effect of low-carb diet. A ketogenic diet was not really tested enough in large clinical trials, but in this six-day study, we can see that ketogenic diet led to reduction in glucose, insulin, the amount of liver fat, but of course, we need longer-term studies to confirm this. Anyway, low-carb diet, like the ketogenic diet, we need to remember that according to the guidelines, this is a short-term phase of the diet. We don't do regularly, on a regular basis, very low-carb diet like the ketogenic diet. It only lasts about eight to 12 weeks, and then gradually, we add foods to the diet, and eventually, in the last phase, we reach a balanced, regular diet following the principles of the Mediterranean diet. So please remember, this is anyway a time-restricted stage of the diet. Is it more beneficial in weight reduction? Yes, but we should use it only in patients with severe obesity who need immediate weight reduction due to comorbidities, and it should be under medical surveillance. And after achieving the goals of the very low-calorie and very low-carb diet, we need to implement long-term lifestyle interventions. And finally, how do we manage to get the patients to be motivated to actually execute a diet? So if they don't understand what NAFLD is, if they don't understand the importance of the diet, they will not have motivation. So our role, as their caregivers, is to explain to them that NAFLD is a treatable disease, the diet can reverse NAFLD features, and then they will have the motivation needed to change their lifestyle. Finally, a few positive words about coffee. Meta-analysis showing that coffee is related with reduced risk of both NAFLD and liver fibrosis. One of the reasons is because coffee has high content of polyphenols. And even better, all coffee types decrease the risk of adverse clinical outcomes related with chronic liver disease. Chronic liver disease in general, chronic liver disease mortality, and ACC, all are reduced by the consumption of coffee. So the takeaways are that weight reduction is very important in the treatment of NAFLD, of course, and the goals for weight reduction are between 5 to 10 percent of the initial body weight, with increasing goals with the presence of NASH and fibrosis. But it's not all about weight reduction. The Mediterranean diet is also effective by itself for the reduction of liver fat. And weight reduction strategies can and should be personally tailored to enable adherence in the long term. This is what is really important, to get adherence in the long term. So let the patients choose their diet they feel most comfortable with. We need to limit sugar, soft drinks, and saturated fat. And that means that we need to reduce the consumption of ultra-processed food, which poses a great challenge in the treatment of NAFLD. For the future, I would be happy to see more large prospective studies with not only steatosis as outcome, but also fibrosis measures as outcome. And also, I would be very happy to elucidate the effect of food processing and preparation methods and its association with NAFLD. And with that, I really thank you for listening. I would like to thank the course organizers, Drs. DeLev and Ranella, for the opportunity to present at this postgraduate course. I'll be talking about the synergism of alcohol and obesity and the effects on cancer. I have a few disclosures. So I'll begin with a discussion of epidemiology and risk factors for alcoholic and non-alcoholic fatty liver disease. Alcoholic liver disease and non-alcoholic fatty liver disease are now the most frequent chronic liver disorders in industrial countries. The advanced forms, alcoholic steatohepatitis and non-alcoholic steatohepatitis are now the most frequent conditions leading to liver cirrhosis and hepatocellular carcinoma in the United States and in Europe. NAFLD is a hepatic manifestation of the metabolic syndrome. With the pandemic rise of obesity, the incidence of NAFLD is also further increasing. And considering the lifestyle in modern societies, there's a significant overlap of risk factors causing NAFLD with alcohol consumption, which also predisposes to alcoholic liver disease in Western countries. Epidemiologic studies propose a causative link between chronic alcohol consumption and progressive liver disease, particularly in obese individuals. Consequently, there are combined additive and synergistic pathological effects of alcohol and obesity on fatty acid levels, respectively on hepatocellular lipid accumulation and injury, as well as hepatic inflammation, fibrosis, and carcinogenesis. Even moderate doses of alcohol have been shown to exhibit synergistic pathological effects with obesity. And this indicates significant differences in the dose threshold for hepatotoxic alcohol effects in lean versus obese subjects. Thus, the, in quotes, safe level of alcohol consumption is substantially lower in obese persons. Interestingly, when we think about alcohol and obesity, it appears as though there are different pathways to the inflammatory liver disease that we see. Typically, for individuals who are using alcohol actively, we see fatty liver, we see inflammation, but that's associated with a high AST and a lower ALT. This, however, primarily reflects the fact that there's active alcohol consumption, as the AST has a shorter half-life than ALT. Thus, with active alcohol consumption and injury, we see a higher AST. On the other hand, in individuals who have obesity-related fatty liver disease, this, of course, is a stable injury to the liver. So we see fatty liver, inflammation, and the resulting injury causes a high ALT and a lower AST because all of the metabolic syndrome factors are associated with a higher ALT. There have been a number of studies that show a synergism between alcohol and obesity. Perhaps one of the best is this study in the Scottish Male Prospective Study of Liver Morbidity and Mortality. Approximately 9,600 men in both main and collaborative studies were studied. They were organized into three groups of BMI groups or classified less than 25, 25 to less than 25, 25 to less than 30 and greater than or equal to 30 kilograms per meter squared BMI in three alcohol groups, none 1 to 14 or 15 or greater units per week. So we had nine groups total, and the outcomes that were observed were liver disease, morbidity, and mortality. Overall, they were shown that there was a relative excess risk due to the interaction between BMI and alcohol of 5.58 and a synergy index of 2.89. So here you can see the baseline risk of one. Baseline plus increase in BMI was associated with a 1.29 risk. Baseline plus alcohol was associated with a 3.66 world increased risk. And the interactions here shows that baseline plus BMI plus alcohol showed us this excess risk due to the interaction of 5.58. Another study which is quite important was done in Taiwan and showed the interaction of between alcohol and obesity in HTC risk in Taiwanese men with chronic HBV infection. Now it's notable that most of the time epidemiologically when we study cohorts of individuals for risk of liver cancer, if we find that they have viral hepatitis, for example, we tend to attribute their risk for cancer to the viral hepatitis and discount contributions from alcohol, obesity, or other risk factors. And this study was really important because it accounted for and looked at the contributions of alcohol and obesity to liver cancer risk in these men. So we had 2,260 Taiwanese men with hepatitis B surface antigen positive in the revealed HBV study cohort. Their mean age was 46, mean BMI was 24, and approximately 20% reported alcohol use. 2.7% had cirrhosis at baseline, and 135 HTCs developed over about 42,000 person years of follow-up. Compared to non-users of alcohol who had a BMI less than 30, alcohol users with a BMI greater than or equal to 30 had a 3.1 times higher relative risk, suggestive of a multiplicative effect. So you can see here the relative risk of individuals with BMI less than 30 who did not drink. If the patients with a BMI less than 30 drank, then they had a 1.8 times increased risk. Individuals with BMI greater than 30 who did not drink had essentially the same risk as those with a low BMI, but those who were drinkers then had a 3.1 times higher relative risk. This graph shows the incidence curves of HTC development in the individuals, and you can see here that in those individuals who were greater than 30 kilograms per meter squared in BMI and who were alcohol users, the hazard ratio was 3.4, a significantly increased likelihood of development of HTC in this group. Interestingly, the incidence of HTC increases with both the degree of obesity and the duration of alcohol use. So if we look here at normal weight individuals with increasing years of alcohol use, we see relatively stable risk, but for overweight individuals, obese individuals, or extremely obese individuals, increasing durations of alcohol use were associated with substantially higher risks of hepatocellular carcinoma. Interestingly also, the incidence of HTC was shown to increase with degree of obesity only in alcohol users. So here we see non-drinkers versus those who are drinkers, and we see that here's degree of obesity in quartiles, and you can see that the incidence of HTC is increasing in alcohol users, but not in those who are non-drinkers. So a substantial degree of interaction there. So this describes what has been called the Bermuda Triangle for the liver, that the combination of alcohol, obesity, and viral hepatitis combined synergistically to substantially increase risk of liver disease and liver cancer. Now, it's difficult to dissect other different contributions of alcohol and obesity to metabolic-associated fatty liver disease, inflammation, cirrhosis, and cancer. The risks for development of advanced liver disease and cancer are likely multilevel. It is probable that genetic factors separately associate with obesity, diabetes, and hypertension, but all potentially contributes to the insulin-resistant phenotype. And it's also likely that separate risk factors associated with liver inflammation, development of fibrosis in response to inflammation, and progression to liver cancer in the context of inflammation have separate contributions to any individual person's risk of cancer. Particularly with non-alcoholic fatty liver disease, some individuals progress to cancer development without cirrhosis. This is not as well recognized for alcohol-induced liver disease, perhaps because lesser degrees of alcohol-induced liver injury short of cirrhosis may not be recognized in lean persons. So this study shows that alcohol and obesity show synergistic effects on the incidence of HTC even after adjustment for viral hepatitis. So in this study, we have adjustment for viral hepatitis, and what we see here is that individuals who have BMIs less than 30 without alcohol use have essentially the same rate of development of HTC as those who have BMI less than 30 and alcohol use. However, if they have BMI greater than 30, that's associated with an increase in risk of hepatocellular carcinoma, and if they have BMI greater than 30 as well as alcohol use, we have a substantially greater risk for hepatocellular carcinoma. So I'll talk now about the pathogenesis of alcoholic and non-alcoholic fatty liver disease. So as I've mentioned, obesity and alcohol use may synergistically worsen hepatic insulin resistance and necroinflammation that can lead to progressive liver injury causing steatohepatitis, which can lead to cirrhosis, and in a subset of these patients, to hepatocellular carcinoma. Interesting additional pathogenic mechanisms include the fact that endotoxin from intestinal bacteria may provoke the transition from steatosis to steatohepatitis in individuals with alcoholic liver disease. Further, alcohol ingestion may sensitize the liver to endotoxin-mediated liver injury, and obese women with fatty livers have been shown to be more sensitive to lipopolysaccharide-induced liver injury. Metformin has been shown to reverse hepatic steatosis and normalize liver enzymes in OB, OB, genetically obese mice. This has been shown to be associated with selective inhibition of hepatic TNF alpha gene expression and the downstream UCP2 and SRBP-1C genes. Fatty acid synthesis, which is an SRBP-1C-activated gene, was also decreased. Molecular mechanisms that may underlie this interaction include cytoprotein 2E1 upregulation, nuclear factor NF-kappa light chain enhancer of activated B-cell activation, cytokine imbalance, particularly upregulation of interleukin-6, activation of STAT3 by IL-6, and lipid peroxidation, all of which can worsen liver injury leading to cirrhosis and HTC. Environmental and other non-genetic factors may amplify the genetic risk that's conferred by different single nucleotide polymorphisms. For example, adiposity has been shown to heighten the risk of liver disease that's associated with polymorphisms in the PNPLA3, TM6SF2, and GCKR genes. There's also a significant effect of concurrent alcohol use, which has provoked a recent suggestion to consider alcoholic fatty liver disease and non-alcoholic fatty liver disease as a single entity within the same spectrum. Although some risk polymorphisms are shared between alcoholic fatty liver disease and non-alcoholic fatty liver disease, others are not. For example, there are polymorphisms in the alcohol and acetaldehyde dehydrogenase genes that uniquely confer a risk of alcoholic liver disease, but not of non-alcoholic fatty liver disease. There are also sex differences that exist in the prevalence of Nafld and liver fibrosis severity, which are higher in men than women during the reproductive age, but change after menopause. Further, the frequency of hepatic steatosis has been shown to vary significantly with ethnicity. For example, in the United States, it's 45% in Hispanics, 33% in whites, 24% in blacks. So there are clearly differences there as well. So alcohol and obesity may have different target levels of genetic susceptibility in progression to inflammation and cirrhosis. So we think about obesity and alcohol, both causing fatty liver. They induce ER stress, mitochondrial dysfunction, and hepatocyte cell death, and subsequently inflammation. But this is modified by the presence of diabetes, by particular gut microbiota, by high ALT and genetic susceptibility. And these are, as we've mentioned, may be different for alcoholic liver disease than for fatty liver disease. This then, of course, is integrated with risk, for example, of cirrhosis due to things like telomerase alterations, which may be associated with a family history of cirrhosis, to fibrosis, and on to cirrhosis, followed by hepatocellular carcinoma. Very briefly, some of the cirrhosis mechanisms particular to NASH include microbiome-derived signals, visceral adipose and insulin-resistant-derived signals, and inflammatory signals. And these combine to induce steatosis and hepatocytes, mediation by DAMS, osteopontin, and lipid peroxidation, leading to activation of immune cells, which secretes cytokines and cause activation of stellate cells. There are also other fibrogenic signals that come directly from steatotic hepatocytes to activate stellate cells. In the case of alcoholic cirrhosis, we know that alcohol use creates acetaldehyde production, and that leads to development of an NLRP3 inflammasome that's associated with immune cell activation. We see also a number of inflammatory cell signals that are induced, leading to activated stellate cells. So what are the clinical implications of this interaction between alcoholic and non-alcoholic fatty liver disease? One is that it can be difficult to distinguish whether there's an alcoholic component from a non-alcoholic basis from steatohepatitis. And this ALD slash NAFLD index can be helpful in distinguishing the two and is a helpful tool to determine whether there's a substantial component of alcohol in patients with steatohepatitis. In summary, the presence of metabolic syndrome and excessive alcohol use have both been independently associated with the risk of death in patients with hepatic steatosis. In a Finnish study in patients with liver steatosis, alcohol use was associated with a dose-dependent increase in risk of advanced liver disease and cancer. In a South Korean study in patients who progressed to severe NAFLD, low levels of alcohol consumption were associated with increased risk of development of steatosis and advanced fibrosis in both obese and non-obese individuals. So even low levels of alcohol consumption were significant. The conclusion is that patients with NAFLD should be strongly advised to completely abstain from alcohol. The safe limits of alcohol use may not exist. Potential therapies that can be chemopreventive in this setting and reduce the risk of liver cancer include weight reduction, reduction alcohol consumption. There's been discussion about esadenosylmethanin, which has shown no difference in patients with chronic alcoholic liver disease. It's not clear whether it might have a beneficial effect in individuals who are both obese and have chronic alcoholic liver disease. It's possible that sterilizing the gut using, for example, rifaximin for selective intestinal decontamination may improve thrombocytopenia and alcoholic cirrhosis and may also reduce the risk of liver cancer. And there's lots of studies investigating whether probiotics then may improve neutrophil responses, cytokine response, and liver function in this context. There are a number of FXR agonists in trials for both NASH and alcoholic cerebral hepatitis. And it's perhaps the case that the SGLT2 inhibitors, which have been shown to substantially decrease obesity, may also be preventive in this setting. So key takeaways are that alcohol and obesity combine synergistically to induce liver injury, cirrhosis, and hepatocellular carcinoma. No amount of alcohol use appears to be without risk, particularly in the context of obesity. Increasing degrees of obesity and increasing duration of alcohol use are associated with progressively increased risk of cirrhosis and hepatocellular carcinoma. And obesity and alcohol likely contribute substantially to risk of cirrhosis and HTC in persons with chronic viral hepatitis, but their contributions are frequently not recognized or appreciated. There are also important inherited genetic components to risk of cirrhosis and HTC that operate at multiple levels of the pathophysiological cascade. And given their current status as the major causes of advanced liver disease and HTC in most industrialized countries, as well as their contributions to risk of HTC in the context of chronic viral hepatitis, it is important to institute broad public health measures to mitigate their effects. Health promotion programs and policies that limit the rise of obesity and alcohol use are needed, similar to measures that are currently used to curb smoking. Finally, mechanistic studies of the underlying biological pathophysiological mechanisms driving the effects of obesity and alcohol use on liver injury are helping to define targets for chemoprevention of liver injury and liver cancer, such as the potential use of metformin, aspirin, or statins. Thanks very much for your attention. Hi, everybody. I would like to start by thanking the organizers for inviting me at this postgraduate course for the liver meeting. I will talk today about the genetic and environmental modifiers of disease severity in non-alcoholic fatty liver disease. This is about me, my experience, what I've done so far. This is my disclosure session. Now we can start. First of all, what common gene variants increase the susceptibility to non-alcoholic fatty liver disease? This is a natural history of the disease, as you know, from normal level to steatosis, inflammation, and cirrhosis, and hepatocellular carcinoma. On the left corner, you see what are the causes of fatty liver. As you know, excess in body weight is the main risk factor, but also the quality of food is implicated as for in high carbohydrate, and especially fructose is deleterious for fatty liver disease, or a diet that is low in polyunsaturated fatty acids and high in saturated fatty acids. On the right side, just pop up the common genetic variants that increase or protect against the progression of fatty liver disease. The orange one, PMPLA3, TM6, and so on, are the one that increases the risk of the disease, while the one color code in blue, HSD17B13, are the one in which the minor allele or the mutant allele is associated with a protection against the disease. I will show you now how these genes interact with the environment, and I will start by talking mostly about PMPLA3, and then I will show you data where all these SNPs are pulled in a genetic risk score, and to show the interaction with obesity in the UK biobank. PMPLA3 associates with liver fat content, and this is the slide identifying the genetic variants. As you see, these are the p-values for the association of fat content in the DALSAR study, and there's only one genetic variant that exceed Bonferroni correction, and you see it above the dashed red line. This is the RS738409 that encodes for a nucleotide variation that turns into a aminoacidic change in the PMPLA3 protein, in which the mutant, the isoleucine, becomes a methionine at position 148M. On the right corner, you see the hepatic fat content stratified by genotype, and you see that carrying the genetic variant increases the liver fat content. What is the frequency of this gene variant in L2 Europeans in homozygosity? It's only five percent. But what is really striking is the enrichment of homozygosity across the entire spectra of fatty liver disease, in which when you think of hepatocellular carcinoma at the top on NASH, nearly one in two Europeans are homozygous for the PMPLA3 genetic variant. This gives extremely large odds ratio for just for liver disease and for cancer that are greater than 12-fold. The frequency of the variant also is different in different ethnic groups, in which in African Americans is lower than Europeans and that it increases in Asians and the highest is in Hispanics. Just to show you some data still from the United States longitudinal data in which they examine the liver disease mortality across the three different genotypes, you see that homozygous for the PMPLA3 genetic variant have a more than eight-fold increased risk of having a exodus from liver disease, and this is in the general population. You may appreciate that the pattern of disease is nearly recessive because you have some effect when you are heterozygous, but the strongest effect, more than additive, I'd say multiplicative is present in homozygous. This is an example of the interaction. On the left side, you see the hepatic fat fraction stratified by the three genotypes in which the C stands for the I isoleucine and the G for the methionine, that's the nucleotide change. That's stratified across the PMPLA3 genotype. On the x-axis, you see the total carbohydrate intake. This study examined a total of 153 Hispanics. As you see, it's from the population. You see how the proportion of mutant PMPLA3 is even higher than the wild type, it's reversed, so the minor allele here is actually the major in Europeans. Here you clearly see that in homozygotes for the PMPLA3 genetic variant, there is an increase in the hepatic fat content with the carbohydrate intake, while in the other two genotypes, there is a decrease and that gives interaction indicating that carbohydrates are not good for carrier of the PMPLA3. This is pretty much the same with a total sugar intake that is on the right side where there's even a stronger interaction. But this is no wonder because PMPLA3 is upregulated by SRBP1C that in turn is upregulated by insulin. No wonder why the effect of the genetic variant is higher the more carbohydrate you get, and that's because you produce more of the bad protein. These are data from the DALASAR study in which you see still the hepatic fat content stratified by different BMI classes. You see that across overweight and then first-degree obesity and severe obesity, you have that the effect of the genetic variant is stronger. The larger the BMI, the higher the BMI, the stronger the effect. Now, this is the genetic risk score pulling in all these genetic variation except HST17B13, that does not affect liver fat content in the UK Biobank where PDFF, so hepatic fat content was available. You see that the more allelic variant you have, and the higher is the slope of the interaction between BMI and the variant in determining fat. You see that the green is lower than the orange and then the top green above is even higher, when you have all these genetic variants and the interaction of this is highly significant. Meaning that it's not only PMPLA3 interacting with this, but generally speaking, all genetic variants, protective or not, have a larger effect size in protection or increasing the susceptibility, the higher the BMI is. What this interaction means when it comes to response to treatment? Well, this has an important effect. Now, I will show you some data here. A major implication of gene-environment interaction is that lifestyle or drug-induced body weight reduction that it could also be given by GLP-1 agonist, for example, will have a different effect in those that are genetically driven. For example, here you see a study in which they induced quite modest body weight, about three kilos in wild-type for the PMPLA3 and homozygous mutant for the genetic variation. You see on the right side of the screen that the reduction in those carrying the bad allele is much larger than those wild-type so you nearly have a 50 percent reduction in liver fat content in those as compared to maybe 20 percent, 15, 20 percent reduction when you do not have the genetic variation. This could be also the case if you were giving GLP-1 agonist, for example, to show also to think of the variability that you get when you give a drug. In the case of NAFLD, these genetic variants will modulate the effect of the response treatment of this drug. Another example is this Wellcome trial in which omega-3 fatty acids were given to a number of individuals to test reduction in circulating triglycerides, but also liver fat content. Then authors, Chris Byrne at University of Southampton, genotyped also the PMPLA3 and what they showed was if you see where the red circle is, that while in carriers of the wild-type variant, there was a reduction in the liver fat content after this omega-3 purified omega-3 administration. Actually, in carriers in homozygotes for the mutant PMPLA3, 148M, there was actually an opposite trend. It was actually slightly increased. This showed how a genetic variant can really modify or to some extent reverse the effect of a potential treatment. Promise of gene-targeted therapies. Among the genes that I showed you, PMPLA3, but also HSD17B13 that were identified by these genetic studies, genome-wide association study, are also promising targets to treat NAFLD. I will start by PMPLA3. I've introduced to you the genetic variant. Now, this is what we think is the mechanism by which this is working. PMPLA3 is on the lipid droplets membrane. Together with other lipases like ATGL, so as a hydrolase activity against triglycerides upon Qs. ATGL is also a co-factor, that it's CGI58. PMPLA3 doesn't have it. When you have the mutant protein, the bad protein doesn't work, doesn't have hydrolase activity, and also sequestrate, so gets within the co-factor of ATGL, impeding the hydrolysis and the turnover of triglycerides in lipid droplets. The idea here is that if we get rid of this bad protein, we will restore the system potentially by allowing ATGL and CGI58 to do their job and reduce liver fat content. We tested this idea in mice, in which we gave a NASH inducing diet for 26 weeks. We also inhibited the endogenous protein, the wild type, but also created a knocking, so mutant PMPLA3 knocking protein, and we inhibit it by antisense oligonucleotide. We evaluate, among other traits, liver steatosis, inflammation, and fibrosis. This is what I will show you now. First, let me show you the beautiful reduction at a protein level on lipid droplets of PMPLA3. This was the same for the wild type. I'll show you only the data in the mutant. We had a really good downregulation of protein synthesis, and also a good reduction. Yeah, I think it was really good reduction in the triglyceride and lipid fat content as measured here by oligodontic staining or by enzymatic assays. Downregulation of PMPLA3 was good for steatosis in the mutant. Here you see the scoring system. But also, it was good in reducing inflammation and fibrosis. If when we were looking at the wild type, actually, the downregulation only reduced the steatosis and not inflammation or steatosis. You see here a little summary. Now, let's go to HSC17B13. This is the genome-wide association study identified this genetic variant that was associated at the very beginning with ALT. They started by doing GWAS on ALT. This is a much more recent discovery. It's only three, four years that that was out. Then you see in red the genetic variant that was identified that was exceeding Bonferroni, sorry, multiple test adjustment, and so it was highly significant. Now, I just look in the literature to meta-analysis just to show you how consistent is the association between HSC17RS72613567 and protection. This genetic variant, the manual allele, is actually good. It protects against liver cirrhosis. This is a meta-analysis of several studies, but has an even stronger effect in protecting against hepatocellular carcinoma. The data show that it protects against fibrosis and also cancer. There's no effect on steatosis, as I just mentioned before. Now, the genetic variation is an insertion of a nucleotide, and this causes a frame shift with a premature truncation in the protein. So, this single nucleotide change insertion actually is extremely deleterious for the HSC17. As you see here, this is a Western blot, so protein levels of HSC17 in human samples stratified by the three genotypes. You see on the very left that there is a nice protein in the wild type. When you look at the homozygous, you see that the protein is nearly gone. The heterozygous is actually two bands, and the lower one is the one for the truncated form. So, the genetic variation induces a truncated form that is not stable. Authors also look at the enzymatic activity of the protein that was completely gone when you were looking at the genetic variant, and this is IsoD. Isoform D is the mutant form we're talking about now. So, low stability, no enzymatic activity. So, we could say that this is a complete loss of function mutation. What is nice is also that this HSC17B13 belongs to a large family of enzymes, but this specific one is only expressed in the liver. So, it makes it a really attractive candidate if you want to try to avoid off-target organs due to therapy. And because the variation is a loss of function, then downregulation or degradation of this HSC17 would be beneficial for fatty liver disease. And we already have some proof of concept in humans in which in five volunteers with fatty liver disease, it was administered an RNA inhibitor, and then it was done a biopsy after 71 days. So, the average reduction of the protein synthesis, and this is why they did a biopsy, was about more than 80% after the dosing, the dose of the RNA interference. And authors found a more than 48% average reduction in carriers after treatment with this genetic, with the interference RNA against the HSC17B13. And on the top right, the bottom right, you can see the single changes in ALT after the administration. And you see that pretty much, except for one, there is a reduction in all the individuals, in ALT levels in all individuals. Now, right, so, I mean, here we are at the edge of facing precision medicine. If this is true and, you know, the data I showed you in mice turns out to be true in human, and, you know, larger studies show that downregulation of HSC is effective. And the reason is that here, we will have to first genotype the individuals with fatty liver disease to identify homozygotes for PMPLA3 allele to be treated with a specific PMPLA3 inhibitor. And also for HSC17B13, you will get the best benefit in the wild type carrying the protein that is stable and still works, because the prediction is that in etherozygotes, you will have less of an effect, and actually in homozygotes for the variant, you will have no effect because the protein is not working. So you will not have any effect of its downregulation. And with this, I want to thank you for the invitation and I hope you enjoyed the talk. These are the take-home message. Yeah, so there's a lot of genetic variation exacerbated by lifestyle and an interaction between genetic variants and lifestyle. PMPLA3 and HSC17B13 downregulation seems to be great targets to treat fatty liver. I thank you again. Thank you.

alcohol consumption thresholds

metabolic syndrome

liver disease

dietary interventions

Mediterranean diet

sugar consumption

NAFLD

intermittent fasting

low-carb diets

genetic modifiers

PMPLA3

HSC17B13