FIB-3 vs FIB-4 in MASLD: Can Removing Age Improve Liver Fibrosis Screening?
Could FIB-3 reduce age-related misclassification in MASLD fibrosis screening? Explore the 2026 study, diagnostic accuracy, trade-offs and clinical limitations.
Two patients, one fibrosis score—and a problem with age
A 32-year-old patient with metabolic dysfunction-associated steatotic liver disease (MASLD) has metabolic risk factors and may have clinically relevant hepatic fibrosis. Yet a low Fibrosis-4 (FIB-4) score could provide reassurance that is partly influenced by age rather than fibrosis severity.
Consider another patient, aged 72 years, with similar biochemical findings. Because age contributes directly to the FIB-4 calculation, the older patient may receive a substantially higher score and be referred for further testing.
Both scenarios illustrate a recognized limitation of FIB-4: its diagnostic behaviour changes at the extremes of age.
FIB-4 remains an established first-line tool for noninvasive fibrosis risk assessment because it is inexpensive, accessible and calculated from routine laboratory results. However, age-dependent misclassification complicates its interpretation in younger and older adults.
A study published in Clinical Gastroenterology and Hepatology on 10 September 2026 examined whether modifying the influence of age—or removing it entirely—could improve fibrosis screening in MASLD.
The investigators compared five strategies in 938 patients and evaluated their findings in an additional US population cohort.
Their central observation was that removing age from the formula produced a simplified score, FIB-3, with similar overall discrimination and potentially more balanced sensitivity and specificity at the extremes of age.
The findings are clinically relevant, but they do not yet justify replacing FIB-4 in routine care.
Why FIB-4 behaves differently in younger and older adults
FIB-4 combines four variables: age, aspartate aminotransferase (AST), alanine aminotransferase (ALT) and platelet count.
The formula incorporates age directly:
FIB-4 = (Age × AST) / (Platelet count × √ALT)
Consequently, two patients with identical aminotransferase levels and platelet counts can have different FIB-4 values solely because of their ages.
This creates a challenge when the score is used to identify patients who may require additional fibrosis assessment.
In younger adults, the age component can contribute to lower FIB-4 values, potentially reducing sensitivity. In older adults, increasing age can elevate the score independently of fibrosis severity, potentially reducing specificity.
The latter may increase referrals for confirmatory testing, whereas the former may contribute to missed clinically important fibrosis.
Age-adjusted thresholds have been introduced to address part of this problem.
The updated 2026 American Gastroenterological Association (AGA) MASLD care pathway recommends FIB-4 as an initial risk-stratification tool, with a threshold of 1.3 for adults younger than 65 years and 2.0 for those aged 65 or older.
Patients meeting or exceeding the relevant threshold generally require second-tier evaluation, commonly with vibration-controlled transient elastography.
However, changing a cutoff does not remove age from the underlying score.
This distinction motivated the investigators to evaluate alternative approaches.
Five strategies for addressing the age effect
Jeevan Nammi and colleagues, with Fernando Bril as corresponding author, conducted a cross-sectional diagnostic accuracy study using two independent MASLD cohorts.
The first included 604 patients evaluated in a tertiary hepatology setting between 2018 and 2025. Magnetic resonance elastography (MRE) served as the reference standard.
The second included 334 patients from centres in Gainesville, Florida, and Birmingham, Alabama. Liver biopsy was used to determine fibrosis stage.
Across both cohorts, mean age was 54 years, 50% of participants were women, mean body mass index was 34.7 kg/m², and 42% had diabetes.
The investigators compared five methods:
Standard FIB-4: A cutoff of 1.3 regardless of age.
Age-adjusted FIB-4: A cutoff of 2.0 for patients aged 65 years or older, retaining 1.3 below that age.
FIB-3: Removal of age from the conventional FIB-4 calculation.
Capped FIB-4: Substitution of age 35 for patients younger than 35 and age 65 for those older than 65.
Log-transformed FIB-4: Replacement of age with its natural logarithm.
The objective was to compare detection of clinically significant fibrosis (F2 or higher), advanced fibrosis (F3 or higher) and cirrhosis.
The investigators additionally assessed the strategies in 6,930 participants from the 2017–2020 National Health and Nutrition Examination Survey (NHANES).
Cutoffs for FIB-3 and log-transformed FIB-4 were derived in the MRE cohort and applied unchanged to the biopsy and NHANES populations.
This provided an opportunity to assess whether the approaches retained performance beyond the cohort used to establish their thresholds.
Similar discrimination does not mean identical screening performance
Across the 938 patients in the two principal cohorts, 20% had clinically significant fibrosis, 13% had advanced fibrosis and approximately 3% had cirrhosis.
All five strategies demonstrated similar overall discrimination when compared with MRE or liver biopsy.
The area under the receiver operating characteristic curve (AUROC) ranged from approximately 0.76 to 0.78 against MRE and 0.77 to 0.81 against biopsy for clinically significant or advanced fibrosis.
AUROC measures how effectively a test distinguishes patients with a target condition from those without it across possible decision thresholds.
An AUROC of 0.5 indicates discrimination no better than chance, whereas 1.0 represents perfect discrimination.
The similar AUROCs suggest that changing or removing age did not materially reduce the overall ability of these scores to rank patients according to fibrosis status.
However, AUROC does not tell clinicians how a test behaves at the threshold used to decide whether further investigation is required.
That decision depends on sensitivity and specificity.
Two scores may have nearly identical AUROCs while identifying different numbers of patients as screen-positive or screen-negative.
This distinction became particularly important when the investigators examined FIB-3.
FIB-3 changes the sensitivity–specificity balance
Although overall discrimination was similar, FIB-3 generally achieved greater sensitivity than standard FIB-4 at the evaluated thresholds, accompanied by lower specificity.
For detection of clinically significant fibrosis:
FIB-3 demonstrated higher sensitivity but lower specificity than standard FIB-4 for detecting clinically significant fibrosis across both validation cohorts.
In the magnetic resonance elastography (MRE) cohort, FIB-3 achieved a sensitivity of 73.6%, compared with 62.3% for standard FIB-4. However, its specificity was slightly lower at 69.7% versus 74.9%, respectively.
A similar pattern was observed in the liver biopsy cohort, where FIB-3 demonstrated greater sensitivity (83.8% vs 76.3%) but lower specificity (45.7% vs 60.6%) than standard FIB-4.
These findings suggest that FIB-3 may identify a greater proportion of patients with clinically significant fibrosis, potentially reducing missed cases. However, its lower specificity could result in more false-positive classifications and additional diagnostic evaluations.
The clinical trade-off is therefore clear: FIB-3 prioritizes sensitivity, while standard FIB-4 offers greater specificity at the evaluated thresholds.
In the MRE cohort, removing age increased sensitivity by 11.3 percentage points while reducing specificity by 5.2 percentage points.
In the biopsy cohort, sensitivity increased by 7.5 percentage points, but specificity decreased by 14.9 percentage points.
This is not an unqualified improvement.
Higher sensitivity could reduce false-negative results, which is important when the purpose of screening is to identify patients who need further assessment.
Lower specificity, however, could increase positive screens among patients without the target degree of fibrosis, potentially generating additional investigations.
Across strategies, positive predictive values ranged from 32.7% to 40.6%, while negative predictive values ranged from 87.1% to 92.5%.
The modest positive predictive values reinforce the role of these indices as initial risk-stratification tools rather than definitive diagnostic tests.
A positive screen still requires appropriate confirmatory assessment.
The strongest clinical signal emerged at the extremes of age
The most interesting findings came from patients younger than 35 years and those aged 65 years or older.
Among patients younger than 35, FIB-3 identified 66.7% of those with clinically significant fibrosis, compared with 33.3% for standard FIB-4.
FIB-3 specificity in this subgroup was 82.9%.
This suggests a potential advantage in younger adults, among whom conventional FIB-4 may underestimate risk because age lowers the calculated score.
However, only nine patients younger than 35 had clinically significant fibrosis in the pooled analysis.
In patients aged 65 years or older, FIB-3 demonstrated specificity of 50.0%, compared with 35.4% for standard FIB-4.
Sensitivity was 80.0% with FIB-3 and 84.4% with standard FIB-4.
Thus, removing age appeared to reduce false-positive screening results among older adults, with a relatively modest sensitivity difference in the observed cohort.
Yet only 45 patients aged 65 or older had clinically significant fibrosis.
The investigators did not conduct formal statistical comparisons within these small subgroups.
Accordingly, these findings should be described as exploratory rather than definitive evidence of superior age-specific performance.
There is also an important distinction between an age-independent mathematical formula and an age-independent clinical test.
Removing age from the calculation eliminates its direct mathematical contribution. It does not prove that the remaining biological variables—or the score's diagnostic performance—are unaffected by aging.
Larger age-stratified validation studies are needed before claiming that FIB-3 reliably resolves age-related misclassification.
Raising the threshold in older adults creates its own trade-off
One of the study's most clinically relevant comparisons concerned increasing the conventional FIB-4 cutoff from 1.3 to 2.0 in adults aged 65 years or older.
In the pooled analysis, this adjustment markedly improved specificity.
Specificity increased from 35.4% to 79.2%.
However, sensitivity decreased from 84.4% to 37.8%.
In other words, the higher threshold classified more patients without clinically significant fibrosis as screen-negative, but also identified a smaller proportion of those who actually had clinically significant fibrosis.
This observation raises a practical screening dilemma.
Where access to elastography is limited, reducing false-positive results may help avoid unnecessary referrals and testing.
However, increasing specificity at the expense of sensitivity risks missing patients who could benefit from additional fibrosis assessment.
The authors noted that using a single FIB-4 threshold of 1.3 could be reasonable in some settings, while acknowledging circumstances where the higher threshold might remain useful.
These are interpretations of the study findings, not changes to current guidelines.
The existing AGA pathway continues to provide the established framework for clinical use.
What did the population-based analysis contribute?
The investigators also examined performance in 6,930 NHANES participants.
Unlike the two principal MASLD cohorts, this analysis used an unselected population sample.
FIB-3 demonstrated an AUROC of 0.71, compared with 0.67 for standard FIB-4, with a statistically significant difference (P < .001).
The investigators also observed a more balanced sensitivity–specificity profile at younger and older ages.
These findings provide supportive evidence that the approach may retain usefulness outside specialist referral populations.
However, the reference standard was different.
The NHANES analysis used a FibroScan-AST (FAST) score of at least 0.35 rather than histological staging or MRE-confirmed fibrosis.
FAST incorporates AST, which is also a component of the fibrosis indices being evaluated.
This shared variable creates potential for correlated measurements and may overestimate diagnostic performance.
Furthermore, FAST is a surrogate marker related to fibrotic metabolic dysfunction-associated steatohepatitis rather than direct histological verification of fibrosis stage.
Therefore, the NHANES findings are complementary evidence, not definitive independent validation against a fibrosis reference standard.
Why the results are promising but not ready for routine implementation
The study has several strengths.
It compared multiple strategies using identical clinical questions, incorporated independent MASLD cohorts, and evaluated performance against both MRE and liver biopsy.
The investigators also used cutoffs derived in one cohort and applied them to other populations.
Nevertheless, several methodological considerations limit immediate adoption.
The principal cohorts included patients selected for specialist fibrosis assessment. Such populations may differ from patients undergoing screening in primary care, endocrinology or community clinics.
The cross-sectional design assessed diagnostic classification at a particular point in time. It did not determine whether any strategy improved long-term liver outcomes.
The relatively small numbers of patients with clinically significant fibrosis at the extremes of age restricted the precision of subgroup findings.
Only 16 participants in each primary cohort had cirrhosis, making those analyses exploratory.
The study also did not establish how many additional elastography examinations, referrals or missed fibrosis cases would occur if FIB-3 replaced FIB-4 across an entire clinical screening programme.
These downstream consequences matter because screening performance must be judged by more than AUROC alone.
A score that identifies additional patients with fibrosis may be advantageous, but its implementation also needs to be feasible, reproducible and clinically beneficial.
At present, the results support further validation rather than a change in standard practice.
Clinical Takeaway: Simpler scoring is promising, but validation comes first
Removing age from FIB-4 produced a score that maintained overall fibrosis discrimination and showed potentially more balanced sensitivity and specificity among younger and older adults with MASLD.
However, FIB-3 did not consistently improve every diagnostic performance measure. Increased sensitivity was accompanied by lower specificity in the primary cohorts, and the most promising age-specific results were based on small subgroups.
The next research priority is prospective validation in larger, diverse primary-care populations using appropriate fibrosis reference standards and clinically meaningful downstream outcomes.
Future studies should determine whether FIB-3 can reduce missed fibrosis, unnecessary confirmatory investigations or age-related variation in referral decisions without compromising patient safety.
GastroAGI perspective: FIB-3 is a promising candidate for simplifying MASLD fibrosis risk assessment, but it remains an investigational alternative. Guideline-recommended FIB-4 and established second-tier testing should remain the basis of routine clinical decision-making until stronger prospective evidence is available.
The central question is not simply whether removing age makes the equation easier. It is whether doing so makes fibrosis screening more accurate, equitable and clinically useful across the full adult age spectrum.
Primary Reference
Nammi J, Southall W, Huynh K, et al. Assessing Strategies to Mitigate the Effect of Age When Using Fibrosis-4 Index to Screen for Metabolic Dysfunction-Associated Steatotic Liver Disease-Related Liver Fibrosis. Clinical Gastroenterology and Hepatology. Published online 10 September 2026.
DOI: https://doi.org/10.1016/j.cgh.2026.09.003
Supporting clinical news source
Cipriano J. Dropping age from FIB-4 may simplify MASLD fibrosis screening. GI & Hepatology News. 14 October 2026.
Clinical guidance context
American Gastroenterological Association. Updated MASLD Clinical Care Pathway, 2026.
https://news.gastro.org/issues/2026/march/aga-updates-masld-care-pathway/
References
We are pioneers in clinical intelligence, dedicated to helping gastroenterologists harness the power of artificial intelligence to drive precision, efficiency, and patient growth.