Noninvasive identification of probable fibrotic nonalcoholic steatohepatitis across the spectrum of glucose tolerance in the United States.

Aim: Identifying patients with fibrotic nonalcoholic steatohepatitis (NASH) is crucial in order to refer them to specialist care as fibrotic NASH represents one of the major inclusion criteria for clinical trials. The aim of this study is to report the prevalence of fibrotic NASH in the general US population across the spectrum of glucose tolerance and evaluate the performance of the recently proposed Fibrotic NASH Index (FNI).

Methods: This is a cross-sectional study of US adults participating in the 2017-2020 cycles of the National Health and Nutrition Examination Survey. Participants with available data to calculate FNI (which is based on AST, HbA1c and HDL-cholesterol) and with a reliable vibration controlled transient elastography examination were included. We excluded participants with chronic viral hepatitis, significant alcohol consumption or other forms of liver disease. Probable fibrotic NASH was defined as a Fibroscan-AST (FAST) score ≥ 0.35.

Results: We included a total of 6268 participants. The overall prevalence of probable fibrotic NASH was 5.9 % (95 % CI 5.2-6.7) and it increased progressively from participants with normal glucose tolerance (3.7 %, 95 % CI 2-9-4.7) to those with diabetes (14.7 %, 95 % CI 12.1-17.8). The performance of FNI for probable fibrotic NASH was satisfactory in the overall population (area under the receiver operating characteristic curve (AUROC): 0.93, 95 % CI 0.92-0.94) and it maintained a good accuracy also in participants with diabetes (n = 1113, AUROC 0.89, 95 % CI 0.86-0.92). In all groups it outperformed Fibrosis-4.

Conclusions: FNI is an easy and reliable test to screen for NASH and its performance is maintained in patients with diabetes, a condition that was shown to negatively influence the performance of several non-invasive scores.