Association of the fibrosis-4 index with prevalent diabetes among adults: a cross-sectional analysis of NHANES 2003-2018 and a Chinese hospital-based health examination population.

Authors

Zeng Chen, Xiaoqing Chen, Jianqiang Wang, Sheng Hu

Published in

Frontiers in endocrinology. Volume 17. Pages 1881759. Epub Jul 30, 2026.

Abstract

The fibrosis-4 (FIB-4) index incorporates age, aminotransferases, and platelet count and is widely used for liver fibrosis risk stratification. Because age is part of its formula and strongly associated with diabetes, apparent associations between FIB-4 and prevalent diabetes may reflect age-related confounding rather than an independent hepatic signal.
We conducted a cross-sectional analysis of 35,213 adults from NHANES 2003-2018 and 32,456 adults from a hospital-based health examination population in Hainan, China (2022-2025). Survey-weighted and conventional multivariable logistic regression evaluated FIB-4 continuously and by tertiles. Additional analyses assessed nonlinear associations, age parameterization, collinearity, liver-related exclusions, age-free liver indices, and incremental model performance.
In age-group-adjusted NHANES models, each 1-unit increase in FIB-4 was associated with higher odds of prevalent diabetes (OR, 1.13; 95% CI, 1.06-1.22). After replacing age categories with continuous age, this association was attenuated and no longer significant (OR, 1.03; 95% CI, 0.95-1.12; P = 0.518). In the hospital-based population, where Model 3 already used continuous age, the association persisted (OR, 1.36; 95% CI, 1.25-1.48). Collinearity diagnostics did not indicate problematic multicollinearity. Adding FIB-4 to traditional diabetes predictors did not improve NHANES discrimination (AUC, 0.822 vs 0.822) and produced only minimal improvement in the hospital-based population (AUC, 0.756 vs 0.758).
The association between FIB-4 and prevalent diabetes was sensitive to age modeling and differed across datasets. FIB-4 provided little incremental model performance beyond traditional diabetes predictors. These cross-sectional findings do not support FIB-4 as an independent screening, diagnostic, or predictive tool for diabetes.

PMID:
42597398
Bibliographic data and abstract were imported from PubMed on 14 Aug 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.