Epidemiologic Profile and Predictors of Fatty Liver: A Hospital-Based Study

Abstract

Background: Non-alcoholic fatty liver is the most common cause of chronic liver disease with increasing prevalence globally. Settings and Design: The current study is an analytical case control study; conducted in ultrasonography outpatient clinic of Cairo University Hospital. Materials and Methods: 150 consented fatty liver cases and 564 controls were screened for fatty liver infiltration using abdominal ultrasonography. Receiver Operating Characteristics (ROC) curve analysis was performed to explore the discriminant ability of the developed model. Results: Among cases, Age, sex and residence matching contributes 32.7%, 36% and 31.3% mild, moderate and severe degree of fatty liver respectively. Cases showed significantly higher body mass index (BMI), waist circumference (WC), total cholesterol, triglyceride, low density lipoprotein (LDL), and lower high density lipoprotein (HDL) than controls. Cases demonstrated higher prevalence of hypertension (11.3% vs 8.3% respectively), and significantly higher prevalence of diabetes (22% vs. 9.2%) (p=0.03). Severe fatty liver cases were significantly older and had significantly higher WC, BMI, significantly higher association with diabetes mellitus, significantly higher levels of total cholesterol, triglycerides and LDL than non-severe degree cases. The significant predictors of sever fatty liver were BMI, total cholesterol and LDL (P = <0.001, R2 = 0.543). Conclusion: The developed regression equation expressed good validation and calibration. It utilizes an algorithm that can quickly and easily address patients with fatty liver. It would useful as a fast, inexpensive primary screening tool for severe fatty liver.