Non-alcoholic fatty liver disease as an independent manifestation of the metabolic syndrome: Results of a US national survey in three ethnic groups

• Published in: Journal of gastroenterology and hepatology Vol. 28; no. 4; pp. 664 - 670
• Main Authors: Smits, Mark M, Ioannou, George N, Boyko, Edward J, Utzschneider, Kristina M
• Format: Journal Article
• Language: English
• Published: Australia Blackwell Publishing Ltd 01.04.2013
• Subjects: Adult; Cross-Sectional Studies; factor analysis; Factor Analysis, Statistical; Fatty Liver - diagnostic imaging; Fatty Liver - etiology; Female; Health Surveys; Humans; insulin resistance; Male; Metabolic Syndrome - complications; Metabolic Syndrome - diagnostic imaging; Middle Aged; NAFLD; Non-alcoholic Fatty Liver Disease; Nutrition Surveys; obesity; Prevalence; Regression Analysis; Sensitivity and Specificity; Ultrasonography; United States - epidemiology
• ISSN: 0815-9319; 1440-1746; 1440-1746
• DOI: 10.1111/jgh.12106

Background and Aim The metabolic syndrome (MetS) and each of its components are strongly associated with non‐alcoholic fatty liver disease (NAFLD). This has led many investigators to suggest that NAFLD is an independent component of the MetS. We formally tested this hypothesis using confirmatory factor analysis, which allows comparison of different models, with or without including NAFLD as a component of the MetS. Methods We analyzed data from 3846 subjects of the Third National Health and Nutrition Examination Survey (1988–1994). NAFLD was defined by increased liver fat measured by ultrasonography. Results MetS by Adult Treatment Panel III criteria was present in 20.5%, and 30.2% had NAFLD, defined as mild, moderate, or severe ultrasonographic steatosis. Using confirmatory factor analysis, a basic model representing the MetS using its currently accepted components (glucose, waist, triglyceride/high‐density lipoprotein ratio, and mean arterial pressure) showed excellent goodness‐of‐fit statistics. Addition of NAFLD to the model as a fifth independent variable decreased model fit, suggesting that NAFLD is not an additional independent component of the MetS. Analysis by ethnicity showed that addition of NAFLD decreased model fit in Whites but resulted in minor improvements in non‐Hispanic Blacks and Mexican Americans. Conclusions The MetS is strongly associated with NAFLD. However, we found no evidence that NAFLD is an independent component or manifestation of the MetS. Interestingly, ethnic differences might be important in this relationship and require further study.