NAFLD Fibrosis Score

About NAFLD Fibrosis Score

The NAFLD Fibrosis Score (NFS), published by Angulo and colleagues in Hepatology in 2007, is a non-invasive scoring system specifically designed to estimate the probability of advanced liver fibrosis (stage F3-F4 on the Brunt scale) in patients with non-alcoholic fatty liver disease (NAFLD). The score uses seven routinely available clinical and laboratory variables: age, BMI, hyperglycemia (fasting glucose ≥110 mg/dL or diabetes), AST/ALT ratio, platelet count, and albumin. The NFS was derived from a multicenter cohort of 733 NAFLD patients with histologically confirmed fibrosis staging and validated in a separate cohort of 541 patients. The score has been endorsed by the AASLD, EASL, and other professional societies as a first-line fibrosis assessment in primary care for patients with NAFLD.

What the calculator measures

The NFS estimates the probability that a patient with NAFLD has advanced fibrosis (F3-F4). The score is interpreted using two thresholds:

• ≤ -1.455: Low probability of advanced fibrosis (negative predictive value ~88%).
• -1.455 to 0.675: Indeterminate, requires further evaluation with elastography (FibroScan) or referral to hepatology.
• ≥ 0.676: High probability of advanced fibrosis (positive predictive value ~82%); requires hepatology referral.

How the formula works

The NFS formula is: NFS = -1.675 + 0.037 × age + 0.094 × BMI + 1.13 × (hyperglycemia: 1 if yes, 0 if no) + 0.99 × (AST/ALT ratio) - 0.013 × platelets - 0.66 × albumin. Each component reflects a known risk factor for advanced fibrosis in NAFLD:

• Age: Fibrosis progression is age-dependent.
• BMI: Obesity is a risk factor for fibrosis progression.
• Hyperglycemia/diabetes: Diabetes is one of the strongest risk factors for NAFLD progression.
• AST/ALT ratio: AST tends to rise more than ALT in advanced fibrosis (AST/ALT >1 is a marker of cirrhosis).
• Platelet count: Declines as portal hypertension develops.
• Albumin: Declines in advanced liver disease.

The formula was derived using multivariate logistic regression in the derivation cohort. The thresholds (-1.455 and 0.676) were chosen to maximize diagnostic accuracy while maintaining an acceptable indeterminate rate (approximately 25-30% of patients fall in the indeterminate range).

Interpretation

The NFS is interpreted in conjunction with FIB-4 (which has similar accuracy but fewer inputs). The 2023 AASLD NAFLD guidance recommends FIB-4 as the first-line assessment in primary care, with NFS as an acceptable alternative or adjunct. Both scores have similar negative predictive values (~88%) for excluding advanced fibrosis, but the NFS has more indeterminate results. For patients in the indeterminate range, further evaluation with elastography (FibroScan) is recommended. FibroScan liver stiffness <8 kPa excludes advanced fibrosis with high accuracy; liver stiffness ≥10-12 kPa suggests advanced fibrosis; liver stiffness ≥15 kPa suggests cirrhosis. For patients with FibroScan indeterminate results, additional testing with the Enhanced Liver Fibrosis panel or MRE may be considered, or liver biopsy for definitive diagnosis.

Clinical use

The NFS is most useful in primary care NAFLD screening. The score can be calculated from routine clinical and laboratory data (no specialized testing required), making it suitable for primary care settings. The 2023 AASLD NAFLD guidance recommends NFS or FIB-4 as the first-line fibrosis assessment in primary care for all patients with suspected NAFLD. Patients with low NFS (≤ -1.455) can be managed with lifestyle modification (5-10% weight loss, Mediterranean diet, exercise) without hepatology referral. Patients with indeterminate or high NFS require further evaluation with elastography or hepatology referral.

Limitations

The NFS has several limitations. First, it has a relatively high indeterminate rate (25-30% of patients fall in the indeterminate range), requiring further evaluation. Second, the NFS has reduced accuracy in patients without diabetes (the hyperglycemia term is the strongest predictor, and non-diabetic patients have lower NFS values on average). Third, the NFS has reduced accuracy in patients under age 35 (where the age numerator is small) and in patients over age 65 (where the age numerator is large). Fourth, the NFS was derived in a predominantly white, U.S. cohort and may perform differently in other populations. Fifth, the NFS is a probability estimate, not a definitive diagnosis; the gold standard for fibrosis staging is liver biopsy. Sixth, the NFS does not distinguish F3 from F4 (advanced fibrosis from cirrhosis), additional testing is needed for this distinction.

Comparison with FIB-4

FIB-4 (Sterling 2006) and NFS (Angulo 2007) have similar accuracy for detecting advanced fibrosis in NAFLD (AUC ≈ 0.80 for both). FIB-4 has fewer inputs (4 vs. 7) and a lower indeterminate rate (20-25% vs. 25-30%). NFS has the advantage of including BMI and diabetes, which are important risk factors for NAFLD progression. The 2023 AASLD NAFLD guidance recommends FIB-4 as the preferred first-line assessment, with NFS as an acceptable alternative.

How to use the result

A NFS of 0.8 in a 55-year-old with BMI 32, type 2 diabetes, AST/ALT ratio 0.8, platelets 200, and albumin 4.0 indicates high probability of advanced fibrosis. The patient should be referred to hepatology for further evaluation with elastography (FibroScan). If FibroScan shows liver stiffness ≥10 kPa, the patient should be evaluated for liver biopsy and consideration of treatment (e.g. resmetirom for MASH with F2-F3 fibrosis, approved by the FDA in 2024). If cirrhosis is confirmed (F4), the patient requires ongoing hepatology surveillance (HCC screening with ultrasound ± AFP every 6 months, variceal screening with upper endoscopy).