Relationship between albumin-corrected anion gap and non-alcoholic fatty liver disease varied in different waist circumference groups: a cross-sectional study.

ABSTRACT

OBJECTIVES: To investigate the association of albumin-corrected anion gap (ACAG) with non-alcoholic fatty liver disease (NAFLD) and clinically significant fibrosis (CSF) defined by vibration-controlled transient elastography measurements. METHODS: This cross-sectional study including 4531 participants was conducted using the data from the NHANES database of cycles 2017-2018. The outcomes were set as NAFLD vs. non-NAFLD and NAFLD with CSF vs. NAFLD without CSF. The generalized additive model and restricted cubic spline analyses were used to assess the nonlinear relationship. The generalized linear models, logistic regression models, sensitivity analysis, P trend test, subgroup analysis, and mediation analysis were employed to analyze the association. Finally, an ACAG-based model was constructed and evaluated. RESULTS: A higher ACAG level was an independent risk factor for NAFLD (P < 0.05), but not for CSF (P > 0.05). The sensitivity analysis and P trend test results substantiated the significantly positive relationship between ACAG and NAFLD (P < 0.05). Interestingly, the obvious connection between ACAG and NAFLD varied in different waist circumference groups and played a central role in the central obesity group. In addition, alanine aminotransferase and waist circumference were the mediators in their relationship. Moreover, the ACAG-based model performed well in predicting NAFLD. CONCLUSIONS: ACAG level is independently associated with NAFLD but not CSF. ACAG might be a novel and reliable biomarker for predicting NAFLD clinically especially in the central obesity population.