# Longitudinal Examination of the UA/HDL-C Ratio as a Biomarker for Fatty Liver Disease: Findings from a Five-Year Follow-Up. Genetics of Atherosclerotic Disease (GEA) Study

## TL;DR

A study found that the UA/HDL-C ratio can predict fatty liver disease over five years, offering a low-cost alternative for early diagnosis.

## Contribution

The study validates the UA/HDL-C ratio as a long-term, non-invasive biomarker for fatty liver disease.

## Key findings

- FLD patients had significantly higher UA/HDL-C levels than controls at baseline and follow-up.
- Higher UA/HDL-C quartiles correlated with increased FLD prevalence, with over 50% in the highest quartile.
- Logistic regression confirmed a strong independent association between UA/HDL-C and FLD after adjusting for multiple factors.

## Abstract

Background: This longitudinal study assessed the Uric Acid/HDL-Cholesterol (UA/HDL-C) ratio as a prognostic biomarker for fatty liver disease (FLD) during a five-year follow-up of 1022 participants from the Genetics of Atherosclerotic Disease (GEA) Study. FLD is a multifactorial disease associated with cardiometabolic comorbidities, and genetic variants affecting uric acid transport (ABCG2 rs2231142) and hepatic lipid metabolism (PNPLA3 rs738409). Early diagnosis is essential to prevent disease progression; however, standard diagnostics are expensive and not widely accessible, highlighting the need for noninvasive tools. Objectives: The study aimed to validate the UA/HDL-C as a long-term predictor for FLD and its effectiveness in risk stratification, including adjustment for cardiometabolic factors and genetics. Methods: Non-contrast computed tomography was used to diagnose FLD and rs738409 and rs2231142 were genotyped by real-time PCR. ROC curves, Kaplan–Meier survival analysis, and logistic regression were used. Results: The findings show that FLD patients exhibited significantly higher UA/HDL-C than controls at both baseline and follow-up (p < 0.0001). Higher UA/HDL-C quartiles were associated with greater FLD prevalence, exceeding 50% in the highest quartile. The index cut-off points were 0.18 in men and 0.09 in women. ROC analysis showed significant discrimination for FLD (AUC: 0.637 overall, 0.650 in men, 0.626 in women). Conclusions: Logistic regression confirmed a strong independent association between UA/HDL-C and FLD over five years, even after adjustment for genetic, biochemical, and anthropometric factors, OR = 3.53, 95% CI: 2.39–4.68, p < 0.0001. Results suggest this ratio could be an alternative to find and follow FLD early on, especially in places with few resources.