AASLD Abstracts we attempt to describe the correlation between hepatic steatosis and single nucleotide polymorphisms (SNP) in 3 candidate genes- PNPLA3, IL- 6 and IL-28 in patients with non- 3 genotypes CHC infection. Methods We analyzed the role of SNPs rs738409 in PNPLA3, rs12980275 in Il28B and rs2069837 in IL 6 in patients with CHC who have undergone liver biopsy (LB) as part of treatment evaluation between January 2013 and December 2014. DNA extraction was done from LB specimen and SNPs identified and characterized in triplicate. Associations of alleles with the presence and/or severity of hepatic steatosis were evaluated by univariate and multivariate logistic regression analysis, considering all relevant covariates. Metabolic syndrome was defined by the updated NCEP ATP III guidelines. Results Of the total 512 patients, mean age was 44.8 ± 11.4 years (86.6% males), predomi- nantly genotype-4 (66.1%), followed by genotype-1 (11.9%). Pretreatment LB showed (Scheuer classification) stage-0 fibrosis in 24.6 %, stage-1 in 34.6%, stage-2 in 22.9%, stage- 3 in 12.4% and stage-4 in 5.6 % patients. Obesity (Asia-Pacific definition) was noted in 74.7% and metabolic syndrome in 33.1%. On univariate analysis, BMI (p=0.03), waist circumference (p=0.03), HDL levels (p=0.001), presence of metabolic syndrome (p=0.01) and rs738409 SNP of PNPLA3 (p=0.004) were associated with hepatic steatosis. Of all variables, rs738409 SNP of PNPLA3 was independently associated with presence of hepatic steatosis (odds ratio = 1.98, 95% confidence interval = 1.2-3.8, p =0.04). rs12980275 in Il-28B and rs2069837 in IL-6 did not show a correlation with presence or severity of hepatic steatosis on univariate analysis. There was no correlation between any of the SNPs studied and fibrosis stage in LB, grade of inflammation in LB or treatment response (both pegylated interferon-ribavirin and direct acting antivirals) to CHC. Conclusion In a Middle Eastern cohort of patients with chronic non-3 genotypes CHC infection, rs738409 SNP of PNPLA3 was independently associated with presence of hepatic steatosis. However rs12980275 SNP in Il28B and rs2069837 SNP in IL 6 were not associated with presence and/or severity of hepatic steatosis. The significant co-occurrence of obesity and metabolic syndrome in these patients highlights the need for continued surveillance post sustained viral response for liver and non liver related complications. Su1527 THE EFFECT OF FOOD INSECURITY ON ADVANCED FIBROSIS IN DIABETICS WITH NONALCOHOLIC FATTY LIVER DISEASE Russell E. Rosenblatt, Gaurav Ghosh, Adam Buckholz, Catherine Lucero, Brett Fortune, Arun Jesudian, Sonal Kumar, Zachary Sherman Background and Aims: Nonalcoholic fatty liver disease (NAFLD) is a leading cause of liver disease in developed nations. Outcomes in NAFLD are directly related to the stage of fibrosis, with an increased risk of overall and liver-related mortality with increasing stage. Diabetes, obesity, and age are well-established risk factors of advanced fibrosis in NAFLD, while there has been less investigation into the impact of economic factors. Patients with food insecurity have been found to have increased risk of both obesity and diabetes, related to lack of choice of food. This study aims to investigate the impact of food insecurity on advanced fibrosis in diabetics with NAFLD a nationally-representative survey. Method: The National Health and Nutrition Examination Survey (NHANES), a nationally-representative survey, from 2005-2014 was used. Using a validated algorithm, adult patients with diabetes were included. Patients were excluded if they had viral hepatitis, excessive alcohol consumption, were pregnant, or had transaminase elevation > 500 IU/L. Advanced fibrosis was diagnosed based on the presence of an ALT > 40 and either an elevated NAFLD fibrosis score, FIB-4 score, or AST to platelet ratio index. The degree of food security was evaluated based on a validated survey included in the NHANES survey. Multivariate logistic regression was per- formed to evaluate for the presence of advanced fibrosis, and factors were included into the model if they independently increased the risk of advanced fibrosis. Results: A total of 14,169,411 weighted participants were included in the sample. Most (77.0%) patients had full food security (FFS) while 6.2% had very low food security. Patients with very low food security (VLFS) were younger and more likely to be black and Hispanic. Those with VLFS also had a higher mean body mass index and mean hemoglobin A1c but had no significant differences in transaminase levels. When controlling for age, gender, race, insulin dependence, hemoglobin A1c, body mass index, and duration of diabetes, patients with VLFS had a significantly higher odds of advanced fibrosis (OR 3.34, 95% CI 1.29-8.62, p=0.01). Conclu- sion: In our study using a nationally-representative survey, VLFS more than tripled a diabetic patient's risk of having advanced fibrosis when controlling for other factors that can independently increase the risk of advanced fibrosis. This study highlights the importance of accounting for economic factors when evaluating patients with NAFLD. Multivariate logistic regression for advanced fibrosis *Controlling for age, gender, race, body mass index, diabetes duration, insulin dependence, and hemoglobin a1c. Su1528 A POPULATION-BASED PREVALENCE OF NONALCOHOLIC FATTY LIVER DISEASE ASSOCIATED COMORBIDITIES AND PIOGLITAZONE USE AMONG US ADULTS WITH TYPE 2 DIABETES Phuc Le, Alexander Chaitoff, Arthur McCullough, Michael Rothberg Background: Type 2 Diabetes (T2D) affects >9% of US adults. In addition to macro- and microvascular complications, patients with T2D have a higher risk of advanced liver diseases including nonalcoholic steatohepatitis (NASH), fibrosis, cirrhosis, and hepatocellular carci- noma. Pioglitazone, traditionally used for T2D, has recently been proven effective also for S-1168 AASLD Abstracts improving NASH and fibrosis score. We estimated trends in prevalence of nonalcoholic fatty liver disease (NAFLD) associated comorbidities and pioglitazone use among US adults with T2D. Methods: We conducted a cross-sectional, retrospective analysis using continuous National Health and Nutrition Examination Survey (NHANES) data from 2003-2014. We analyzed trends in 2-year and 4-year periods. Study sample included people aged ≥18 years who had an HbA1C >6.4%, fasting plasma glucose >125 mg/dL, or were told they had diabetes by a doctor. We excluded type 1 diabetes, defined as age <20 years who were treated solely with insulin. We further excluded patients with viral hepatitis, excess alcohol consumption, aspartate aminotransferase (AST) or alanine aminotransferase (ALT)>500, and pregnant women. NAFLD was defined as the hepatic steatosis index (HSI) >36. NAFLD- associated advanced fibrosis was defined as BMI ≥25 kg/m2, ALT>40U/l (men) or >30U/l (women) and either APRI score>1, FIB-4 index>2.67 or NFS score>0.676. NASH cirrhosis was defined as APRI score >2 and either ALT>40U/l (men) or >30U/l (women), alkaline phosphatase >113 U/l, or total bilirubin >1.3mg/dl. Prevalence of pioglitazone use was estimated for the overall T2D population and those having suspected NAFLD. Survey weights were incorporated in all prevalence estimates to account for the complex design of NHANES. We used R 3.4.2 for all analyses. Results: A total of 4,613 patients with T2D were included for the 12-year period. Mean age was 58.9 years, 50.7% was male, and mean BMI was 33.1. The overall prevalence of suspected NAFLD, advanced fibrosis, and NASH cirrhosis was 84.3%, 4.3%, and 0.38%, respectively. From 2003-2014, there were no significant changes in the prevalence of NAFLD, advanced fibrosis or NASH cirrhosis (table). Pioglitazone use peaked in 2005-2006 and then declined sharply (p-value <0.05) (figure). Patients with suspected NAFLD were not more likely to receive treatment with pioglitazone. Conclusion: Most patients with T2D patients had suspected NAFLD. One potentially effective treatment, pioglitazone, has been decreasing in use, regardless of whether patients have NAFLD. Use of pioglitazone should be reconsidered for this population. Prevalence of NAFLD associated comorbidities in adult patients with type 2 diabetes Su1529 SERUM IMMUNOGLOBULIN A LEVELS IN ADULT PATIENTS WITH ALCOHOLIC AND NON-ALCOHOLIC FATTY LIVER DISEASE Evan Elias, Julia Uhanova, Kelly Kaita, Stephen Wong, David Peretz, Susan Cuvelier, Eberhard Renner, Gerald Minuk Background: The immune response to gut-derived endotoxins is thought to contribute to the pathogenesis of alcoholic (ALD) and non-alcoholic fatty liver disease (NAFLD). Immunoglobulin A (IgA) synthesis is an important component of that immune response. Objectives: To document the prevalence of elevated serum IgA (E-IgA) levels in adult ALD and NAFLD patients and determine whether serum IgA levels correlate with the severity and course of the liver disease. Methods: Adult patients attending an urban tertiary care centre were identified from a computerized clinical database. Severity and course of liver disease were assessed by standard liver biochemistry, MELD and Fib-4 scores at initial presentation and last follow-up visit. Results: 175 ALD and 941 NAFLD patients were identified. The mean age ± SD of ALD patients was 61±11 years and NAFLD patients 57±13 years. 65% of ALD patients and 49% of NAFLD patients were male. The prevalence of E- IgA was 67% and 27% respectively. The mean ages and gender distributions of E-IgA ALD and NAFLD patients were similar to those with normal IgA levels (N-IgA) in their respective cohorts. E-IgA ALD and NAFLD patients had significantly more impaired hepatic function (hyperbilirubinemia, prolonged INR and hypoalbuminemia) and higher MELD scores than N-IgA patients. The percent of E-IgA ALD and NAFLD patients with Fib-4 scores suggestive of cirrhosis at baseline was higher in the E-IgA than N-IgA cohorts (ALD: 53% vs 34%, p<0.01 and NAFLD: 25% vs 6.5%, p<0.0001). After median follow-up of 53±23 (ALD) and 43±37 (NAFLD) months, biochemical evidence of hepatic dysfunction remained more impaired and MELD scores significantly higher in E-IgA than N-IgA ALD and NAFLD patients. The percent of pre-cirrhotic patients at baseline who developed cirrhosis (by Fib- 4 testing) during follow-up was similar in E-IgA and N-IgA ALD patients (21% vs 17%, p= 0.62) but significantly higher in E-IgA than N-IgA NAFLD patients (12% vs 3.1%, p<0.0001). Conclusions: Elevated serum IgA levels are common findings in ALD and to a lesser extent in NAFLD patients. In both conditions, elevated IgA levels correlate with biochemical evidence of disease severity. However, only in NAFLD are elevated levels at baseline associated with subsequent progression to cirrhosis. These findings suggest that IgA may contribute to the pathogenesis of NAFLD but are less like to be of pathophysiologic importance in ALD.