# The association between triglyceride glucose-waist-to-height ratio and non-alcoholic fatty liver disease in adults aged over 60 in the United States: a cross-sectional study

## TL;DR

This study finds that the triglyceride glucose-waist-to-height ratio (TyGWHtR) is a better predictor of non-alcoholic fatty liver disease in older U.S. adults than other common health indicators.

## Contribution

The study introduces TyGWHtR as a novel, cost-effective indicator for NAFLD screening in older adults.

## Key findings

- TyGWHtR showed the best predictive performance with an AUC of 0.7641 for NAFLD detection.
- TyGWHtR demonstrated a nonlinear dose-response relationship with NAFLD.
- TyGWHtR outperformed five other indicators in predicting NAFLD risk in older adults.

## Abstract

To identify the optimal surrogate indicator for the initial public health screening of non-alcoholic fatty liver disease (NAFLD) among U.S. adults aged 60 years and above, by systematically comparing the predictive performance of a novel index—the triglyceride glucose-waist-to-height ratio (TyGWHtR)—against 15 other conventional obesity, metabolic, and inflammatory indices.

This cross-sectional study included 1,634 participants from the National Health and Nutrition Examination Survey (2017-March 2020). This study included 16 indicators: weight, waist-to-height ratio (WHtR), body mass index (BMI), waist circumference (WC), weight-adjusted-waist index (WWI), triglyceride glucose (TyG), TyGBMI, TyGWHtR, TyGWWI, TyGWC, TyGWeight, cardiometabolic index (CMI), systemic inflammatory response index (SIRI), systemic immune inflammation index (SII), visceral adiposity index (VAI), and lipid accumulation product (LAP). The participants were categorized into a non-NAFLD group (n = 945) and an NAFLD group (n = 689). The associations between 16 potential indicators and NAFLD, as defined by FibroScan®, were evaluated using weighted multivariable logistic regression. Indicators that displayed significant associations were further analyzed using receiver operating characteristic (ROC) curves. To investigate a dose-response relationship, we employed restricted cubic splines. Additionally, subgroup and sensitivity analyses were conducted to assess the robustness of the findings.

Five indicators (WHtR, TyG, TyGWHtR, WWI, and CMI) showed significant positive associations with NAFLD. Among these, TyGWHtR demonstrated the best predictive performance [area under the curve (AUC): 0.7641; 95% CI: 0.7414–0.7869], with a nonlinear dose-response relationship. The association remained robust across all subgroups and sensitivity analyses.

The TyGWHtR index is a simple, low-cost, and effective single indicator that outperforms multiple conventional indices for identifying older U.S. adults at high risk of NAFLD. Its reliance on routinely measured clinical parameters makes it an efficient tool for initial risk stratification in community and public health settings, facilitating targeted prevention and supporting healthy aging initiatives.