Are Non-invasive Tools Less Accurate in NAFLD Patients With Type 2 Diabetes?

Recent literature on non-alcoholic fatty liver disease (NAFLD) has sought to characterise its bidirectional association with type 2 diabetes mellitus (T2DM).
PUBLISHED IN: Clinical Gastroenterology and Hepatology (May 2023)


Advanced liver fibrosis is frequent in patients with T2DM, just as the latter’s presence is known to exacerbate the development of cirrhosis and hepatocellular carcinoma in those with NAFLD. Due to the intricate relationship between both conditions, the European Association for the Study of the Liver, the European Association for the Study of Diabetes, and the American Diabetes Association recommend liver fibrosis evaluation in patients with T2DM. In light of this, corroborating the diagnostic accuracy of non-invasive tools (NITs) for NAFLD detection in patients with T2DM is imperative.

This study aimed to compare the diagnostic accuracy of 6 NITs in a large multicenter cohort of NAFLD patients with or without diabetes. The study endpoint was biopsy-proven advanced fibrosis, and tested tools included vibration-controlled transient elastography (VCTE), NAFLD Fibrosis Score (NFS), FIB-4, FibroTest (FT), FibroMeter-V2G (FM), and FibroMeter-VCTE (FMVCTE).


All NITs tested had significantly lower AUROCs in patients with T2DM than those without. Moreover, at the 2 NIT thresholds used to rule advanced fibrosis out or in, T2DM was associated with a greater reduction in NIT specificity than in sensitivity. For patients with T2DM, specificity at the ruled-in threshold was significantly reduced, generating an increased rate of false positives. Age and heightened T2DM-associated alpha 2-macroglobulin levels were identified as key variables in reducing NIT specificity. Importantly, when used in patients with T2DM, the sequential FIB4-VCTE algorithm reported a diagnostic accuracy of 79.0% compared to its usual 90.3% in patients without T2DM. The FM-FMVCTE and VCTE-FMVCTE were the best-performing algorithms in patients with T2DM, boasting increased sensitivity and a diagnostic accuracy of 85%. These findings highlight the importance of adapting NIT-based diagnostic algorithms to T2DM status in NAFLD patients.

Share the article
Avatar photo

S Duarte, BSc

Articles: 91