American Diabetes Association’s Standard of Care 2023 recommended screening patients with T2D for significant liver fibrosis with FIB-4 and transient elastography (TE). However, the evidence supporting this approach is controversial. The aim of this study was to assess the use of FIB-4, and its correlation with TE, in an unselected population of patients with diabetes. Adult patients from NHANES 2017-2020 with complete data on FIB-4 and TE were included. Diabetes was defined based on FPG, A1c or prior history. Patients with secondary causes of liver steatosis were excluded. NAFLD was defined as a controlled attenuation parameter ≥274 dB/m, while clinically significant and advanced liver fibrosis were defined as TE ≥9.6 and ≥12 kPa, respectively. Among 1193 patients with diabetes (Age: 61 ± 13; BMI: 32.6 ± 7.3 kg/m2; A1c: 7.4 ± 1.7%), the prevalence of NAFLD, clinically significant liver fibrosis, and advanced fibrosis were 68%, 15% and 8%, respectively. These were significantly higher than in patients without diabetes even after stratifying for BMI (all p<0.001). Liver fibrosis worsened with increasing insulin resistance (HOMA-IR), but were not associated with A1c levels. FIB-4 performed poorly to predict significant or advanced liver fibrosis in diabetes (AUC: 0.59 [0.54-0.64] and AUC: 0.63 [0.56-0.69], respectively). Moreover, it performed numerically worse than plasma AST alone (AUC: 0.63, p=0.10 and AUC: 0.68, p=0.10, respectively). Among patients with significant and advanced fibrosis based on TE, FIB-4 was <1.30 (suggested cut-off for screening) in 51% and 47% of patients (i.e., false negatives), suggesting a significant disagreement between these tests. Patients with diabetes are at high risk of liver disease, and therefore, screening for liver fibrosis is warranted. However, our results suggest that current screening strategies based on FIB-4 and TE may not be adequate, as they may miss a significant number of patients with advanced fibrosis.


F.Bril: None. M.Gray: Advisory Panel; Takeda Pharmaceutical Co., Ltd., Theratechnologies, Consultant; Novo Nordisk.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at