Skip to content

Natural history and fibrosis progression

TL;DR — Fibrosis stage, not steatohepatitis, is the histological feature that tracks long-term outcomes: in the NASH CRN cohort no histological feature other than fibrosis was associated with long-term outcomes (Angulo 2015, PMID 25935633), and in a Swedish biopsy cohort followed up to 33 years, overall mortality was not raised in patients with NAS 5–8 but fibrosis F0–2 (HR 1.41, 95% CI 0.97–2.06), while F3–4 raised it regardless of NAS (HR 3.3, 2.27–4.76) (Ekstedt 2015, PMID 25125077; Hagström 2017, PMID 28803953). Liver-related mortality rises roughly exponentially with stage — mortality rate ratio 1.58 at F1, 2.52 at F2, 3.48 at F3, 6.40 at F4 for all-cause, and 9.57, 16.69 and 42.30 at F2, F3 and F4 for liver-related death (Dulai 2017, PMID 28130788). Progression is slow and bidirectional: in paired-biopsy meta-analysis, 33.6% progressed, 43.1% were stable and 22.3% regressed, at 0.07 stages/year from F0 in MASL versus 0.14 in MASH — one stage per 14.3 years and 7.1 years respectively (Singh 2015, PMID 24768810). Progression to cirrhosis occurs in 3–5% of patients and usually takes more than 20 years (Hagström 2024, PMID 39243773). Once decompensation begins the clock changes completely: median survival from first decompensation to death or transplant is 2.0 years (Noureddin 2024, PMID 38571305). The unresolved tension is that steatohepatitis — the target of the approved drugs' primary endpoint — is a poor independent prognostic marker, in part because its histological definition is subjective (Akbari 2024, PMID 38293684).

The central finding: fibrosis, not steatohepatitis

Study Design n Follow-up Finding
Angulo 2015, PMID 25935633 Multinational biopsy cohort 619 median 12.6 y Fibrosis stage, and no other histological feature (steatosis, ballooning, inflammation, NAS), was associated with long-term outcomes
Ekstedt 2015, PMID 25125077 Swedish biopsy cohort 229 mean 26.4 y (range 6–33) NAS did not predict overall mortality; F3–4 did (HR 3.3, 2.27–4.76). Excess mortality from CVD (HR 1.55), HCC (HR 6.55), cirrhosis (HR 3.2), infection (HR 2.71)
Hagström 2017, PMID 28803953 Swedish biopsy cohort 646 up to 40 y Fibrosis stage but not NASH predicted mortality and time to severe liver disease
Akbari 2024, PMID 38293684 Swedish non-cirrhotic cohort vs matched population 1,260 vs 12,529 median 13 y 20-year cumulative incidence of major adverse liver outcomes: 2% in the reference population, 3% at F0, 35% at F3. No difference in risk between patients with and without NASH in the biopsy subcohort

Angulo's hazard ratios for death or liver transplantation, versus F0, were 1.88 (95% CI 1.28–2.77) at F1, 2.89 (1.93–4.33) at F2, 3.76 (2.40–5.89) at F3 and 10.9 (6.06–19.62) at F4; for liver-related events (26 patients, 4.2%) they were 14.2 (3.38–59.68) at F3 and 51.5 (9.87–269.2) at F4. Age (HR 1.07 per year), diabetes (1.61, 1.13–2.30) and current smoking (2.62, 1.67–4.10) also predicted the composite, while statin use was associated with lower risk (HR 0.32, 0.14–0.70) (PMID 25935633).

Hagström converted stage into time, which is the form most useful for counselling. Over a mean 20 years (range 0–40; 139,163 person-years), 12% of NAFLD patients versus 2.2% of matched controls developed severe liver disease (p<0.001); hazard ratios versus controls ranged from 1.9 at F0 to 104.9 at F4, and adding NASH to the model did not improve prediction at any stage (likelihood ratio test p>0.05 throughout). The lower bound of the 95% CI for the 10th percentile of time to severe liver disease was 22–26 years at F0–1, 9.3 years at F2, 2.3 years at F3, and 0.9 years to decompensation at F4 (PMID 28803953).

Akbari's finding is the sharpest statement of the problem: a histological NASH label added nothing prognostically once fibrosis was known, and the authors attribute this explicitly to the subjectivity of the histological definition rather than to biological irrelevance — steatohepatitis is most likely causal for fibrosis progression, but the diagnosis of it carries limited prognostic value. See histology and biopsy.

Stage-specific mortality

Fibrosis stage All-cause MRR (Dulai 2017) Liver-related MRR (Dulai 2017) All-cause HR (Ng 2023) Liver-related HR (Ng 2023)
F0 1 (ref) 1 (ref) 1 (ref) 1 (ref)
F1 1.58 (1.19–2.11) 1.41 (0.17–11.95) — —
F2 2.52 (1.85–3.42) 9.57 (1.67–54.93) 1.46 (1.08–1.98) 4.07 (1.44–11.5)
F3 3.48 (2.51–4.83) 16.69 (2.92–95.36) 1.96 (1.41–2.72) 7.59 (2.80–20.5)
F4 6.40 (4.11–9.95) 42.30 (3.51–510.34) 3.66 (2.65–5.05) 15.1 (5.27–43.4)

Sources: Dulai 2017, PMID 28130788 (5 cohorts, 1,495 patients, 17,452 patient-years); Ng 2023, PMID 35513235 (14 studies, 17,301 patients, reconstructed individual participant data). Taylor 2020 (13 studies, 4,428 patients) gives concordant unadjusted risk ratios for F0 versus F4: all-cause mortality RR 3.42 (2.63–4.46), liver-related mortality RR 11.13 (4.15–29.84), liver transplant RR 5.42 (1.05–27.89), liver-related events RR 12.78 (6.85–23.85) (PMID 32027911).

The confidence intervals on liver-related mortality ratios are enormous because liver deaths are rare in the reference stratum. The direction is unambiguous; the point estimates are not precise, and the three meta-analyses disagree by roughly two- to three-fold at F4 (MRR 6.40 vs HR 3.66 vs RR 3.42), a spread explained by whether estimates are adjusted, by follow-up duration and by the composition of the reference group. Dulai's own stated limitation is that patients with simple steatosis and NASH without fibrosis were pooled into the reference group.

Absolute risk is more useful than ratios. In Ng 2023, all-cause mortality at 1, 5 and 10 years was 0.1%, 3.3% and 7.7% for F0–2, versus 0.3%, 20.6% and 41.5% for F4 (PMID 35513235). In the Global-MASLD cohort of 17,792 biopsy-confirmed patients (mean follow-up 6.6 y), 5-year mortality was 2.1% overall, rising to 8.3% in cirrhosis and exceeding 10% in those with high-risk non-invasive test values (Younossi 2026, PMID 41231627).

Rates of progression — and of regression

Population Annual fibrosis progression Time per stage Source
NAFL (MASL) with F0 at baseline 0.07 stages/yr (0.02–0.11) 14.3 y (9.1–50.0) Singh 2015, PMID 24768810
NASH (MASH) with F0 at baseline 0.14 stages/yr (0.07–0.21) 7.1 y (4.8–14.3) PMID 24768810
Overall trajectory across 411 paired biopsies 33.6% progressed, 43.1% stable, 22.3% regressed over 2,145.5 person-years — PMID 24768810
Any MASLD → cirrhosis 3–5% of patients, usually >20 years — Hagström 2024, PMID 39243773
Fibrosis progression events (incidence) 49.0 per 1,000 person-years — Le 2024, PMID 38281814

Two consequences follow. First, regression is common — roughly one in five paired-biopsy patients improved a stage without any drug — which sets a demanding placebo-response floor for trials and explains why single-arm histological improvement is uninterpretable. Second, the mean rates conceal wide heterogeneity: some patients progress from F0 to cirrhosis within a decade while most never leave F0–F1, and the paired-biopsy literature is enriched for people biopsied twice, which is not a random sample.

What drives the divergence. In 671 NASH CRN participants with ≥2 serial biopsies (444 adults, 227 children), the PNPLA3 rs738409 G allele increased fibrosis-progression risk (aHR 1.31, 95% CI 1.05–1.64) while the HSD17B13 rs72613567 A allele reduced fibrosis progression (aHR 0.69, 0.51–0.92) and increased both MASH resolution (aHR 1.58, 1.13–2.22) and fibrosis regression (aHR 1.42, 1.09–1.85). An unweighted 3-SNP risk score predicted progression (fibrosis aHR 1.37, 1.18–1.60); TM6SF2 rs58542926 did not influence histological change. Type 2 diabetes and BMI change independently influenced histology, and age, sex, BMI and T2D significantly modified the genetic effects (Vilar-Gomez 2026, PMID 40998180). See genetics.

Incidence of adverse events

Pooled incidence rates per 1,000 person-years across 79 cohort studies and 1,377,466 people with NAFLD (Le 2024, PMID 38281814):

Event Rate per 1,000 PY
All-cause mortality 14.6
CVD mortality 4.53
Non-liver cancer mortality 4.53
Liver-related mortality 3.10
Any liver-related event 24.3
Fibrosis progression 49.0
Cirrhosis 10.9
Liver transplant 12.0
Hepatocellular carcinoma 3.39
Incident diabetes 19.0
Incident CVD 24.77
Renal impairment 30.3
Non-liver cancer 10.5

Note the baseline composition — 21.97% already had cirrhosis and 58.85% biopsy-proven NASH — so these are rates in a referral-enriched population, not in unselected MASLD. Biopsy-proven NASH had higher HCC incidence than non-NASH (p=0.043). Time-period analysis showed decreasing CVD and non-liver-cancer mortality alongside increasing decompensated cirrhosis (p<0.05); no differences by sex.

Cause of death: the composition changes with stage

In the Swedish nationwide biopsy cohort (n=10,568 versus 49,925 matched controls, median 14.2 years), overall mortality was 28.6 versus 16.9 per 1,000 person-years (aHR 1.93, 95% CI 1.86–2.00), with a monotonic gradient: simple steatosis aHR 1.71 (1.64–1.79), non-fibrotic NASH 2.14 (1.93–2.38), non-cirrhotic fibrosis 2.44 (2.22–2.69), cirrhosis 3.79 (3.34–4.30). The excess mortality was dominated by extrahepatic cancer (4.5/1,000 PY, aHR 2.16) and cirrhosis (2.7/1,000 PY, aHR 18.15), with cardiovascular disease (1.4/1,000 PY, aHR 1.35) and HCC (1.2/1,000 PY, aHR 11.12) contributing more modestly (Simon 2021, PMID 33037056).

This is where the literature genuinely conflicts. Unselected NAFLD populations show cardiac deaths outnumbering liver deaths roughly 4.5:1 (Younossi 2023, PMID 36626630), but in the biopsy-confirmed Asian CLIONE registry (n=1,398, median 4.6 y), liver-related mortality was the leading cause of death, with liver-specific mortality 2.34/1,000 PY (95% CI 1.52–3.58) against overall mortality 5.34/1,000 PY (4.02–7.08) and HCC incidence 4.17/1,000 PY (3.02–5.75) (Fujii 2023, PMID 35051649). The same registry found fibrosis independently associated with liver-related events but not with overall mortality after adjustment — the opposite of the Swedish result. Ascertainment explains most of this: biopsied cohorts are enriched for advanced disease, Asian cohorts have different competing risks, and follow-up durations differ by a factor of three or more. Applying the MASLD definition to the same CLIONE cohort reclassified 99% of NAFLD cases as MASLD and left prognosis unchanged (Iwaki 2024, PMID 38263684).

The reconciliation is stage-dependent, and should be stated that way: in the general MASLD population cardiovascular death dominates, whereas among those who reach F3–F4 liver-related death becomes the leading cause. See cardiovascular and extrahepatic outcomes.

Does MASLD add risk beyond the metabolic dysfunction that defines it?

This is the most consequential unresolved question about the natural history, because the 2023 definition makes cardiometabolic dysfunction a criterion rather than a confounder — so any comparison of MASLD against a general population is confounded by construction. The one study that removed that confounding by restricting the comparator to people who also have cardiometabolic risk factors reached an uncomfortable conclusion. Among 10,750 NHANES III participants with any cardiometabolic risk factor, followed for up to 26.7 years (Kwak 2025, PMID 38739848):

Group vs "no steatotic liver disease" (all with ≥1 CMRF) All-cause Cardiovascular Cancer Liver
MASLD aHR 1.04 (0.95–1.14), p=0.413 0.88 (0.75–1.04), p=0.139 1.06 (0.84–1.33), p=0.635 —
MetALD aHR 1.41 (1.05–1.89), p=0.022 — 2.35 (1.33–4.16), p=0.004 15.04 (2.96–76.35), p=0.002

The MetALD excess was more pronounced in those with advanced fibrosis by FIB-4. Read literally, hepatic steatosis per se added nothing to all-cause, cardiovascular or cancer mortality once the comparator carried the same metabolic burden; the excess risk in this design belonged to alcohol. Two qualifications matter and neither dissolves the finding: NHANES steatosis was defined by ultrasound rather than histology, so advanced disease is diluted by a large majority of simple steatosis; and the analysis is of mortality, not of liver events, where the stage gradients above are unambiguous. Set against the Swedish biopsy cohort's monotonic mortality gradient from simple steatosis (aHR 1.71) upward (PMID 33037056) — where the comparator was matched general-population controls, not metabolically matched — the two designs answer different questions, and only the Kwak design answers the one clinicians actually face.

MASLD, MetALD and ALD are not the same natural history

The 2023 statement created MetALD as a category; four large analyses have now tested whether it behaves differently, and they agree that it does for the liver and does not for the heart.

Source Design Liver outcomes Cardiovascular / mortality
Celsa 2025, PMID 40953570 meta-analysis, 24 cohorts, 11,575,558 individuals (9,801,312 MASLD, 1,774,246 MetALD) MetALD vs MASLD: liver-related events HR 1.62 (1.16–2.25), p=0.0086; HCC 1.33 (1.00–1.77), p=0.048; extrahepatic cancers 1.03 (1.01–1.06), p<0.0001 (I²=0%) cardiovascular events HR 0.96 (0.85–1.09), p=0.48; all-cause mortality 1.08 (0.97–1.19), p=0.14 — no difference
Miwa 2026, PMID 42120953 UK Biobank, 244,760 participants; 47,949 MASLD (19.6%), 23,000 MetALD (9.4%), 8,085 ALD (3.3%); median 13.5 y MALO per 1,000 PY: MASLD 1.91, MetALD 2.35, ALD 5.80. Adjusted vs MASLD: ALD HR 2.95 (2.64–3.29), MetALD HR 1.23 (1.11–1.36) MACE per 1,000 PY 14.88 / 15.48 / 19.36; mortality 12.99 / 13.59 / 20.93. ALD vs MASLD: MACE HR 1.17 (1.11–1.24), mortality 1.52 (1.44–1.60). MetALD and MASLD did not differ on MACE or mortality
Kwak 2025, PMID 38739848 NHANES III, 26.7 y, comparator restricted to CMRF-positive MetALD liver mortality aHR 15.04 (2.96–76.35) MetALD all-cause 1.41 (1.05–1.89); MASLD null on all endpoints

Three consistent conclusions. Alcohol is the dominant modifier of hepatic outcome within the steatotic-liver spectrum — the MetALD-versus-MASLD liver hazard is 1.23–1.62 and the ALD-versus-MASLD hazard is 2.95, an ordering that recovers the old alcohol/non-alcohol distinction the rename was designed to move past. Cardiovascular risk does not track the alcohol axis at all: MACE and all-cause mortality are indistinguishable between MASLD and MetALD in every dataset. And the heterogeneity in the meta-analysis is severe (I²=76–93% for most outcomes, with only extrahepatic cancer at I²=0%), so the pooled hepatic hazard ratios should be read as directional rather than precise. This is also the strongest existing argument for why the alcohol boundary must be measured objectively rather than asked about (nomenclature and definitions).

After decompensation

Once hepatic decompensation begins, the natural history compresses. In 2,016 well-characterised MASLD patients, 220 (11%) decompensated over a median 3.2 years. The first event was ascites in 69.5%, hepatic encephalopathy in 25% and variceal haemorrhage in 13.6% (8.1% had more than one simultaneously). Of those with a first event, 39.5% had a second at a median of 0.7 years, and 8.1% a third at a median of 1.3 years after the second. Median survival from first decompensation to death or transplant was 2.0 years; 33.1% died and 14% were transplanted (Noureddin 2024, PMID 38571305). See cirrhosis and decompensation.

Non-invasive tests predict outcomes as well as histology

The prognostic argument for biopsy has largely collapsed. In an individual-participant-data meta-analysis of 2,518 biopsy-proven NAFLD patients from 25 studies followed a median 57 months, 5-year time-dependent AUCs for a composite of death, HCC, transplant or cirrhosis complications were 0.72 (95% CI 0.62–0.81) for histology, 0.76 (0.70–0.83) for liver stiffness by VCTE, 0.74 (0.64–0.82) for FIB-4 and 0.70 (0.63–0.80) for NAFLD fibrosis score — all significant after adjustment (Mózes 2023, PMID 37290471). Akbari found biopsy and FIB-4 gave near-identical 10-year C-indices (≈0.73 vs 0.72) (PMID 38293684).

Serial measurement adds more. In 16,603 patients across 16 centres (median follow-up 51.7 months, 316 liver-related events), the VCTE-based Agile 3+ and Agile 4 scores achieved integrated time-dependent AUC 0.89 for liver-related events. Among 10,920 with repeat examination, 81.9% had a stable Agile 3+ risk category; the incidence of liver-related events was 0.6 per 1,000 person-years with persistently low Agile 3+ versus 30.1 per 1,000 person-years with persistently high — a 50-fold separation — and in high-score patients a >20% decrease was associated with substantially reduced risk (Lin 2024, PMID 38512249). See noninvasive assessment.

What predicts progression

From the Global-MASLD cohort (n=17,792, advanced fibrosis ≥F3 in 35%): T2D prevalence rose stepwise with fibrosis stage from 28% at F0 to 70% at F4 (trend p<0.0001). Independent predictors of advanced fibrosis were older age, T2D and obesity — though the obesity association varied by region. Older age, male sex, T2D and obesity independently predicted both mortality and clinical events (all p<0.05) (Younossi 2026, PMID 41231627).

Predictor Effect Source
Type 2 diabetes 28% at F0 → 70% at F4; independent predictor of advanced fibrosis, mortality and events Younossi 2026, PMID 41231627
PNPLA3 G allele fibrosis progression aHR 1.31 (1.05–1.64) Vilar-Gomez 2026, PMID 40998180
HSD17B13 A allele fibrosis progression aHR 0.69 (0.51–0.92); regression aHR 1.42 (1.09–1.85) PMID 40998180
BMI change independently influences histological trajectory PMID 40998180
Baseline steatohepatitis vs steatosis doubles progression rate (0.14 vs 0.07 stages/yr) Singh 2015, PMID 24768810
Persistently high Agile 3+ 30.1 vs 0.6 liver-related events per 1,000 PY Lin 2024, PMID 38512249

Open questions

  • Does MASLD add mortality risk beyond the cardiometabolic dysfunction that defines it? With the comparator restricted to people who also carry a cardiometabolic risk factor, MASLD was null for all-cause (aHR 1.04, 0.95–1.14), cardiovascular and cancer mortality over 26.7 years, while MetALD was not (PMID 38739848). Every other cited mortality cohort compares MASLD against unmatched controls and is therefore confounded by the definition itself. No biopsy-based cohort has replicated the metabolically matched design.
  • Why does alcohol shift hepatic but not cardiovascular risk? MetALD carries 1.23–1.62-fold the liver-event hazard of MASLD across a meta-analysis of 11.6 million people and UK Biobank, with identical MACE and all-cause mortality (PMIDs: 40953570, 42120953). If both conditions share the metabolic driver of cardiovascular disease, the alcohol increment should be liver-specific — which it is — but no study has tested whether the same is true at matched fibrosis stage.
  • Does histological improvement predict outcome improvement? Taylor's systematic review set out to assess whether change in fibrosis stage is a valid surrogate endpoint and concluded that further studies are needed to establish it (PMID 32027911). That conclusion has not been superseded by a completed trial demonstrating treatment-effect transfer to hard hepatic outcomes. Query re-run 2026-09-02: (NAFLD OR NASH OR MASLD OR MASH) AND (fibrosis stage) AND (mortality OR "liver-related events" OR decompensation) AND (cohort OR meta-analysis) — 204 records; no completed primary clinical-outcome treatment trial was identified.
  • Why do biopsy-based cohorts disagree about the leading cause of death? Simon (Sweden) attributes most excess mortality to extrahepatic cancer and cirrhosis with modest CVD contribution (PMID 33037056); Fujii (Asia) finds liver-related death leading and fibrosis unassociated with overall mortality (PMID 35051649); unselected populations show cardiac death dominating (PMID 36626630). No study has decomposed how much of this is ascertainment, region, competing risk and follow-up length.
  • Who are the fast progressors? Mean progression is 0.07–0.14 stages/year, but 3–5% reach cirrhosis. Genetic scores explain part of the variance (PMID 40998180) and are not used clinically. No prospective study has tested whether genotype-informed stratification identifies fast progressors well enough to change surveillance intervals.
  • What does fibrosis regression mean prognostically? 22.3% of paired-biopsy patients regressed a stage spontaneously (PMID 24768810), and a >20% fall in Agile 3+ associated with lower event risk (PMID 38512249). Whether regression restores the risk of the lower stage, or leaves a residual "scarred" risk, is untested.

References

  1. Angulo P, Kleiner DE, Dam-Larsen S, et al. Liver Fibrosis, but No Other Histologic Features, Is Associated With Long-term Outcomes of Patients With Nonalcoholic Fatty Liver Disease. Gastroenterology. 2015;149(2):389-97.e10. PMID 25935633
  2. Ekstedt M, Hagström H, Nasr P, et al. Fibrosis stage is the strongest predictor for disease-specific mortality in NAFLD after up to 33 years of follow-up. Hepatology. 2015;61(5):1547-54. PMID 25125077
  3. Hagström H, Nasr P, Ekstedt M, et al. Fibrosis stage but not NASH predicts mortality and time to development of severe liver disease in biopsy-proven NAFLD. J Hepatol. 2017;67(6):1265-1273. PMID 28803953
  4. Dulai PS, Singh S, Patel J, et al. Increased risk of mortality by fibrosis stage in nonalcoholic fatty liver disease: Systematic review and meta-analysis. Hepatology. 2017;65(5):1557-1565. PMID 28130788
  5. Taylor RS, Taylor RJ, Bayliss S, et al. Association Between Fibrosis Stage and Outcomes of Patients With Nonalcoholic Fatty Liver Disease: A Systematic Review and Meta-Analysis. Gastroenterology. 2020;158(6):1611-1625.e12. PMID 32027911
  6. Ng CH, Lim WH, Hui Lim GE, et al. Mortality Outcomes by Fibrosis Stage in Nonalcoholic Fatty Liver Disease: A Systematic Review and Meta-analysis. Clin Gastroenterol Hepatol. 2023;21(4):931-939.e5. PMID 35513235
  7. Singh S, Allen AM, Wang Z, et al. Fibrosis progression in nonalcoholic fatty liver vs nonalcoholic steatohepatitis: a systematic review and meta-analysis of paired-biopsy studies. Clin Gastroenterol Hepatol. 2015;13(4):643-54.e1-9. PMID 24768810
  8. Hagström H, Shang Y, Hegmar H, et al. Natural history and progression of metabolic dysfunction-associated steatotic liver disease. Lancet Gastroenterol Hepatol. 2024;9(10):944-956. PMID 39243773
  9. Akbari C, Dodd M, Stål P, et al. Long-term major adverse liver outcomes in 1,260 patients with non-cirrhotic NAFLD. JHEP Rep. 2024;6(2):100915. PMID 38293684
  10. Simon TG, Roelstraete B, Khalili H, et al. Mortality in biopsy-confirmed nonalcoholic fatty liver disease: results from a nationwide cohort. Gut. 2021;70(7):1375-1382. PMID 33037056
  11. Fujii H, Iwaki M, Hayashi H, et al. Clinical Outcomes in Biopsy-Proven Nonalcoholic Fatty Liver Disease Patients: A Multicenter Registry-based Cohort Study. Clin Gastroenterol Hepatol. 2023;21(2):370-379. PMID 35051649
  12. Iwaki M, Fujii H, Hayashi H, et al. Prognosis of biopsy-confirmed metabolic dysfunction-associated steatotic liver disease: A sub-analysis of the CLIONE study. Clin Mol Hepatol. 2024;30(2):225-234. PMID 38263684
  13. Le MH, Le DM, Baez TC, et al. Global incidence of adverse clinical events in non-alcoholic fatty liver disease: A systematic review and meta-analysis. Clin Mol Hepatol. 2024;30(2):235-246. PMID 38281814
  14. Noureddin N, Huang DQ, Bettencourt R, et al. Natural history of clinical outcomes and hepatic decompensation in metabolic dysfunction-associated steatotic liver disease. Aliment Pharmacol Ther. 2024;59(12):1521-1526. PMID 38571305
  15. Mózes FE, Lee JA, Vali Y, et al. Performance of non-invasive tests and histology for the prediction of clinical outcomes in patients with non-alcoholic fatty liver disease: an individual participant data meta-analysis. Lancet Gastroenterol Hepatol. 2023;8(8):704-713. PMID 37290471
  16. Lin H, Lee HW, Yip TC, et al. Vibration-Controlled Transient Elastography Scores to Predict Liver-Related Events in Steatotic Liver Disease. JAMA. 2024;331(15):1287-1297. PMID 38512249
  17. Younossi ZM, de Avila L, Petta S, et al. Predictors of fibrosis, clinical events, and mortality in MASLD: Data from the Global-MASLD study. Hepatology. 2026;84(1):204-215. PMID 41231627
  18. Vilar-Gomez E, Yates KP, Kleiner DE, et al. Genetic and non-genetic drivers of histological progression and regression in MASLD. J Hepatol. 2026;84(3):502-516. PMID 40998180
  19. Younossi ZM, Golabi P, Paik JM, et al. The global epidemiology of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH): a systematic review. Hepatology. 2023;77(4):1335-1347. PMID 36626630
  20. Kwak M, Kim HS, Jiang ZG, et al. MASLD/MetALD and mortality in individuals with any cardio-metabolic risk factor: A population-based study with 26.7 years of follow-up. Hepatology. 2025;81(1):228-237. PMID 38739848
  21. Celsa C, Pennisi G, Tulone A, et al. Risk of hepatic and extrahepatic outcomes associated with metabolic dysfunction-associated steatotic liver disease and metabolic dysfunction and alcohol-associated steatotic liver disease: a systematic review and meta-analysis. Lancet Gastroenterol Hepatol. 2025;10(11):998-1012. PMID 40953570
  22. Miwa T, Díaz LA, Harberts A, et al. Subtype Differences in Major Adverse Liver Outcomes and Cardiovascular Events and Mortality in Steatotic Liver Disease: A UK Biobank Analysis. Aliment Pharmacol Ther. 2026;64(3):367-378. PMID 42120953