Prevalence and Determinants of Undiagnosed Liver Steatosis and Fibrosis in a Nationally Representative Sample of US Adults.

Background Chronic liver diseases account for approximately 1.9 million deaths globally every year and negatively affect health-related quality of life. Early detection of liver disease may enable timely treatment, potentially improving patient outcomes. This study aimed to determine the prevalence and determinants of liver steatosis and fibrosis in US adults with no previously diagnosed liver condition. Methods We conducted an observational, nationally representative, cross-sectional study using data from the National Health and Nutrition Examination Survey (NHANES) conducted from January 2017 to March 2020. Study participants were 7,391 adults aged 21 and older with no history of diagnosed liver disorders who underwent vibration-controlled transient elastography (VCTE) to determine liver steatosis and fibrosis. Controlled attenuation parameter (CAP) values between 248 and 267 dB/m were classified as mild steatosis, and those over 267 dB/m as advanced steatosis. Liver stiffness measurement (LSM) values between 7.65 and 13 kPa were classified as moderate/severe fibrosis, and those over 13 kPa as cirrhosis. Covariates included age, sex, race, body mass index (BMI), diabetes mellitus, kidney disease, smoking history, alcohol intake, alanine aminotransferase (ALT), aspartate aminotransferase (AST), physical activity, sedentary time, and sleep time. The associations of subject characteristics with liver CAP and LSM were evaluated using survey multivariable linear regression. Shapley Additive Explanations values determined the relative importance of each attribute in the model. The discriminative performance of classification models was assessed using the area under the receiver operating characteristic (AUROC) curve. Results The population prevalence of liver steatosis was 57.2% (10.2% mild; 47.0% advanced). The relative importance of covariates in predicting liver CAP was 63.1% for BMI, 10.7% for ALT, and less than 10% for the other covariates. The prevalence of significant fibrosis was 11.4% (8.3% moderate/severe fibrosis; 3.1% cirrhosis). The relative importance of covariates in predicting LSM was 67.3% for BMI and less than 10% for the other covariates. BMI alone demonstrated acceptable discriminative performance in classifying varying severities of steatosis and fibrosis (AUROC range 72%-78%) at cutoffs between 28 and 33 kg/m