Skip to main content
Clinical and Molecular Hepatology logoLink to Clinical and Molecular Hepatology
. 2026 Jun 5;32(4):e2026.0496. doi: 10.3350/cmh.2026.0496

Risk assessment of cardiovascular disease-related mortality in metabolic dysfunction-associated steatotic liver disease

Donghee Kim 1,✉, Pojsakorn Danpanichkul 2, Do Han Kim 3, Donghyun Ko 4, Karn Wijarnpreecha 5,6,†, Aijaz Ahmed 1
PMCID: PMC13748841  PMID: 42244098

Metabolic dysfunction-associated steatotic liver disease (MASLD) is currently the most prevalent chronic liver disease in the USA [1]. Growing evidence indicates that MASLD is a major risk factor for cardiovascular disease (CVD) [2], which remains the leading cause of death among individuals with MASLD, responsible for approximately one-third of all fatalities in this population [3,4]. These data support the role of risk assessment for CVD in individuals with MASLD. Improving our understanding of phenotypic variation in MASLD may help inform the development of preventive measures to reduce all-cause and CVD-related mortality in this population [3]. Although clinical prediction scores like the Framingham Risk Score (FRS, 2008) [5], the Pooled Cohort Equation (PCE, 2013) [6], and Predicting Risk of CVD EVENTs (PREVENT) equations [7], are designed to predict 10-year CVD events, their utility for predicting long-term mortality (>10 years) in individuals with MASLD remains uncertain. We investigated the independent association of clinical CVD prediction scores in MASLD with all-cause and cause-specific mortality using the National Health and Nutrition Examination Survey (NHANES) III (1988–1994).

The NHANES III dataset provides a nationally representative sample of non-institutionalized USA adults. Of the 13,856 adults (aged 20–74 years) who underwent hepatic ultrasonographic examinations at a mobile examination center, we excluded 1,248 individuals with missing data on mortality status, body mass index, and laboratory tests, including high-density lipoprotein-cholesterol and triglycerides. Out of 4,846 individuals with steatotic liver disease (SLD), we also excluded 1,039 participants due to significant alcohol intake, viral hepatitis, prolonged use (over 6 months) of medications associated with SLD, pregnancy, and those who did not meet cardiometabolic criteria. We also excluded 332 participants with a history of CVD (congestive heart failure, coronary heart disease, angina, heart attack, or stroke). Subsequently, the final study cohort comprised 3,475 adults with MASLD. This study employed clinical and laboratory methods in accordance with those described in prior studies [8-10]. SLD was identified as any sonographic evidence of hepatic steatosis ranging from mild to severe in the absence of other liver disease, significant alcohol use, or exposure to steatogenic medication. MASLD was classified as the presence of SLD with one or more cardiometabolic criteria [11,12]. We evaluated the FRS, PCE, and PREVENT equations to estimate the 10-year risk of atherosclerotic CVD (ASCVD), and the PREVENT equation to estimate the 10-year risk of total CVD [5-7]. Risk categories were defined as follows: low risk (estimated 10-year risk below 5%); borderline risk (5–7.4%); intermediate risk (7.5–19.9%); and high risk (≥20%). The detailed methodology is shown in Supplementary Methods.

We analyzed 3,475 individuals with MASLD and without CVD (mean age, 46.1 years; 51.9% men). Over the median follow-up period of 26.5 years (interquartile range, 21.4–28.0 years), there were 1,405 deaths (40.4%). CVD accounted for the most common cause of death in 436 individuals (31.0%), while cancer was responsible for 338 of the deaths (24.0%). In Table 1, higher cardiovascular risk scores predicted greater all-cause mortality among individuals with MASLD. Compared to those with a low risk below 5%, individuals with high risk scores had a two- to threefold increased risk of mortality (highest hazard ratio [HR], 3.26 [95% confidence interval (CI), 2.42–4.39] for PREVENT 10-year CVD risk and lowest HR, 2.60 [95% CI, 1.74–3.87] for FRS), even after adjusting for age, sex, race/ethnicity, marital status, education level, smoking status, alcohol consumption, waist circumference, and advanced fibrosis. The risk of mortality also increased incrementally for each percentage or category increase in risk score (highest HR, 1.49 [95% CI, 1.34–1.66] per each category for the PREVENT CVD risk score and 1.05 [95% CI, 1.04–1.07] per 1% increase for the PREVENT ASCVD risk score and lowest HR, 1.38 [95% CI, 1.21–1.58] for FRS and 1.02 for PCE or FRS). Higher cardiovascular risk scores were strongly associated with increased risk of cardiovascular mortality in individuals with MASLD. Compared to individuals with low risk, those with high risk had significantly higher hazards of cardiovascular mortality across all models. Intermediate categories yielded weaker, non-significant associations in the PCE and FRS models, whereas the PREVENT ASCVD and CVD risk scores remained statistically significant, even at the intermediate level. Mortality risk also increased incrementally for each percentage or category increase in risk score (highest HR, 1.07 [95% CI, 1.05–1.10] per 1% increase for the PREVENT ASCVD equations and lowest HR, 1.03 [95% CI, 1.02–1.04] for FRS). When we performed the same analyses among individuals without MASLD and CVD, the overall results were largely similar (Supplementary Table 1). Since aging was associated with a notable increase in cardiovascular risk scores, we explored the relationship between these scores and cancer-related mortality in MASLD. There was no significant, dose-dependent relationship between cardiovascular risk scores and cancer-related mortality (Supplementary Table 2). For predicting all-cause mortality in MASLD (Supplementary Table 3), PREVENT 10-year ASCVD and CVD risk scores achieved the highest point estimates for C-statistics (0.847–0.852), while FRS had the lowest point estimates for C-statistics at 0.827, with substantial overlap in the CIs across all four scores. For cardiovascular mortality, PREVENT scores (0.806) also ranked highest, while FRS consistently showed lower discrimination (0.784), although the CIs overlapped. When we repeated the analysis among individuals without MASLD, the C-statistics were largely similar for cardiovascular risk scores predicting all-cause and cardiovascular mortality.

Table 1.

Association between cardiovascular risk score and all-cause and cardiovascular mortality in individuals with MASLD (n=3,475)

All-cause mortality
Cardiovascular mortality
No. of deaths Age and sex-adjusted model
Multivariable model
No. of deaths Age and sex-adjusted model
Multivariable model
HR (95% CI) HR (95% CI) HR (95% CI) HR (95% CI)
PREVENT 10-year ASCVD risk
 <5% (n=2,226) 445 1 1 115 1 1
 5% to <7.5% (n=363) 228 1.67 (1.29–2.17) 1.61 (1.25–2.08) 60 1.32 (0.76–2.31) 1.25 (0.73–2.16)
 7.5% to <20% (n=810) 667 2.37 (1.86–3.02) 2.04 (1.60–2.61) 239 3.13 (2.08–4.70) 2.59 (1.74–3.85)
 ≥20% (n=76) 65 3.41 (2.12–5.49) 2.79 (1.83–4.24) 22 3.41 (1.41–8.25) 2.61 (1.19–5.72)
 Per 1 category increase 1.52 (1.36–1.70) 1.41 (1.27–1.57) 1.74 (1.44–2.10) 1.58 (1.32–1.89)
 Per 1% increase 1.07 (1.05–1.08) 1.05 (1.04–1.07) 1.09 (1.06–1.11) 1.07 (1.05–1.10)
PREVENT 10-year CVD risk
 <5% (n=1,925) 300 1 1 74 1 1
 5% to <7.5% (n=302) 136 1.53 (1.14–2.06) 1.39 (1.03–1.87) 40 1.71 (0.84–3.47) 1.54 (0.78–3.05)
 7.5% to <20% (n=923) 685 2.51 (1.79–3.50) 2.18 (1.59–2.99) 227 2.77 (1.53–4.99) 2.32 (1.31–4.09)
 ≥20% (n=325) 284 4.05 (3.00–5.48) 3.26 (2.42–4.39) 95 5.75 (3.03–10.94) 4.46 (2.37–8.40)
 Per 1 category increase 1.60 (1.43–1.79) 1.49 (1.34–1.66) 1.79 (1.47–2.18) 1.64 (1.34–2.01)
 Per 1% increase 1.04 (1.03–1.05) 1.03 (1.02–1.04) 1.05 (1.03–1.07) 1.04 (1.03–1.06)
PCE 10-year ASCVD risk
 <5% (n=1,876) 308 1 1 78 1 1
 5% to <7.5% (n=310) 135 1.67 (1.22–2.29) 1.39 (1.05–1.83) 31 2.07 (1.06–4.04) 1.72 (0.90–3.29)
 7.5% to <20% (n=807) 547 2.31 (1.73–3.08) 1.73 (1.29–2.32) 169 2.68 (1.34–5.37) 1.97 (0.99–3.90)
 ≥20% (n=482) 415 3.83 (2.80–5.24) 2.75 (2.02–3.74) 158 7.58 (3.80–15.12) 5.57 (2.65–11.66)
 Per 1 category increase 1.55 (1.40–1.71) 1.39 (1.25–1.54) 1.95 (1.57–2.44) 1.77 (1.38–2.27)
 Per 1% increase 1.03 (1.03–1.04) 1.02 (1.02-1.03) 1.05 (1.04–1.06) 1.04 (1.03–1.05)
FRS
 <5% (n=1,429) 170 1 1 40 1 1
 5% to <7.5% (n=302) 85 1.61 (1.04–2.50) 1.44 (0.94–2.21) 27 2.38 (1.08–5.23) 2.12 (0.96–4.67)
 7.5% to <20% (n=881) 450 1.88 (1.29–2.73) 1.56 (1.07–2.27) 115 1.82 (1.04–3.19) 1.49 (0.82–2.68)
 ≥20% (n=863) 700 3.46 (2.27–5.29) 2.60 (1.74–3.87) 254 4.83 (2.62–8.90) 3.51 (1.91–6.46)
 Per 1 category increase 1.52 (1.33–1.74) 1.38 (1.21–1.58) 1.72 (1.44–2.05) 1.55 (1.30–1.85)
 Per 1% increase 1.02 (1.02–1.03) 1.02 (1.01–1.02) 1.03 (1.02–1.04) 1.03 (1.02–1.04)

The multivariable model was adjusted for age, sex, race/ethnicity, education status, marital status, smoking status, alcohol consumption, waist circumference, and advanced fibrosis. The leading cause of death was classified using the Underlying Cause of Death 113 (UCOD_113) codes, based on the International Classification of Diseases, 10th revision. Cardiovascular mortality was defined by deaths due to diseases of the heart (I00-I09, I11, I13, I20-I51) and cerebrovascular diseases (UCOD_113: 55-68, 70).

MASLD, metabolic dysfunction-associated steatotic liver disease; HR, hazard ratio; CI, confidence interval; PREVENT, Predicting Risk of Cardiovascular Disease Events; ASCVD, atherosclerotic cardiovascular disease; CVD, cardiovascular disease; PCE, pooled cohort equation; FRS, Framingham risk score.

In this nationally representative study, all established cardiovascular risk scores reliably predicted all-cause and cardiovascular mortality in MASLD, whereas the PREVENT equations provided effective risk stratification, especially in higher-score categories associated with substantially greater risk of mortality. In addition, PREVENT equations were effective at stratifying the intermediate-risk categories. Furthermore, the PREVENT scores showed a clear dose-response relationship, with higher scores correlating with progressively greater risk of mortality.

The PREVENT equations are a newly developed and recommended tool for the risk stratification of CVD [7]. Previous studies reported that, after adjusting for clinical and metabolic confounders, MASLD may not be associated with increased all-cause mortality [3,8,9]. The phenotypic characterization of individuals with MASLD by incorporating outcome-predicting parameters can improve clinical management and inform targeted prevention strategies [3,13]. Although both MASLD and CVD are linked to underlying metabolic disturbances, CVD may further increase the risk of mortality in individuals with MASLD who may have adverse underlying metabolic profiles. Utilizing the PREVENT and other cardiovascular risk scores in individuals with MASLD can help identify those at increased risk for both all-cause and cardiovascular mortality. Since CVD remains the leading cause of mortality in individuals with MASLD [3], higher PREVENT scores may serve as essential indicators of CVD-related risk in this population. Although the routine use of PREVENT equations in clinical practice is not yet established, they may be employed in the future to assess the CVD risk among individuals with MASLD. Further studies are needed to confirm these findings.

One of the main strengths of our study is the extensive follow-up (median, 26.5 years), the considerable breadth of demographic, clinical, and metabolic data, and the definition of MASLD using hepatic ultrasonography. There are several limitations in the study. First, because of a small sample size, the NHANES data on liver-related mortality were not available for analysis. Second, a key limitation is the strong age-dependence of all cardiovascular risk scores. The observed dose-response pattern may thus partly reflect the associated underlying age gradient. While our models adjusted for age, this introduces potential for over-adjustment and attenuated interpretability, since age is embedded in each score. However, the scores were not associated with age-related cancer mortality, suggesting they capture risk information beyond the general effect of aging. Third, there is an overlap between the diagnostic criteria of MASLD and the predictors included in the PREVENT equations that warrants caution and consideration. This ‘construct overlap’ may provide a structural advantage for the PREVENT equations in MASLD. While PREVENT had the highest point estimates for the C-statistic, there was substantial overlap in the CIs across all four scores. In addition, the higher C-statistic for the PREVENT scores may simply reflect its richer set of predictors rather than superior discrimination in MASLD. We were unable to perform formal pairwise tests or assess model calibration because these analyses were not feasible using the NHANES survey design after adjusting for appropriate sample weights, stratification, and clustering. Therefore, our finding of PREVENT’s superior discrimination is preliminary and should be interpreted with caution. Fourth, the NHANES lacks longitudinal data on cardiovascular risk scores and MASLD status, restricting our ability to evaluate temporal changes in these variables. Finally, although we attempted to adjust for known risk factors to assess the independent relationship between cardiovascular risk scores and mortality, unmeasured confounders may have influenced the outcomes observed in our study.

In conclusion, this study demonstrates that the new PREVENT equations can effectively stratify the risk of long-term cardiovascular mortality among individuals with MASLD, a population not explicitly represented in the score’s original derivation cohort. While our analysis assesses mortality over an extended follow-up period, it does not constitute formal validation of the scores for their intended 10-year incident-event prediction. The findings support the utility of PREVENT as a valuable prognostic tool for identifying elevated risk of adverse long-term outcomes in MASLD.

Abbreviations

CI

confidence interval

CVD

cardiovascular disease

FRS

Framingham Risk Score

MASLD

metabolic dysfunction-associated steatotic liver disease

NHANES

National Health and Nutrition Examination Survey

PCE

Pooled Cohort Equation

PREVENT

Predicting Risk of CVD EVENTs

SLD

steatotic liver disease

Footnotes

Authors’ contributions

Donghee Kim participated in study concept and design, acquisition of data, data analysis, interpretation of data, drafting of the manuscript, critical revision of the manuscript, and study supervision. Pojsakorn Danpanichkul, Do Han Kim, Donghyun Ko, and Karn Wijarnpreecha participated in data interpretation and critical revision of the manuscript. Aijaz Ahmed participated in study concept and design, interpretation of data, drafting of the manuscript, critical revision of the manuscript, and study supervision.

Conflicts of Interest

The authors have no conflicts to disclose.

SUPPLEMENTARY MATERIAL

Supplementary material is available at Clinical and Molecular Hepatology website (http://www.e-cmh.org).

Supplementary Table 1.

Association between cardiovascular risk score and all-cause/cardiovascular mortality in individuals without MASLD (n=8,408)

Supplementary Table 2.

Association between cardiovascular risk score in MASLD and cancer-related mortality

Supplementary Table 3.

Assessment of cardiovascular risk score to predict all-cause and cardiovascular mortality based on the presence or absence of MASLD

SUPPLEMENTARY METHODS

REFERENCES

  • 1.Kim D, Danpanichkul P, Wijarnpreecha K, Cholankeril G, Loomba R, Ahmed A. Current burden of steatotic liver disease and fibrosis among adults in the United States, 2017-2023. Clin Mol Hepatol. 2025;31:382–393. doi: 10.3350/cmh.2024.0987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Kim D, Danpanichkul P, Wijarnpreecha K, Ahmed A. Burden of cardiovascular complications in steatotic liver disease in the United States. Clin Mol Hepatol. 2026;32:e326–e330. doi: 10.3350/cmh.2026.0249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Konyn P, Ahmed A, Kim D. Causes and risk profiles of mortality among individuals with nonalcoholic fatty liver disease. Clin Mol Hepatol. 2023;29(Suppl):S43–S57. doi: 10.3350/cmh.2022.0351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Kim D, Danpanichkul P, Wijarnpreecha K, Ahmed A. Contemporary trends in extrahepatic mortality of chronic liver disease in the United States from 2014 to 2023. Clin Mol Hepatol. 2026;32:e24–e28. doi: 10.3350/cmh.2025.0802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.D’Agostino RB Sr, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, et al. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008;117:743–753. doi: 10.1161/CIRCULATIONAHA.107.699579. [DOI] [PubMed] [Google Scholar]
  • 6.Goff DC, Jr, Lloyd-Jones DM, Bennett G, Coady S, D’Agostino RB, Gibbons R, et al. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association task force on practice guidelines. Circulation. 2014;129(25 Suppl 2):S49–S73. doi: 10.1161/01.cir.0000437741.48606.98. [DOI] [PubMed] [Google Scholar]
  • 7.Khan SS, Matsushita K, Sang Y, Ballew SH, Grams ME, Surapaneni A, et al. Development and validation of the American Heart Association’s PREVENT equations. Circulation. 2024;149:430–449. doi: 10.1161/CIRCULATIONAHA.123.067626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Kim D, Kim WR, Kim HJ, Therneau TM. Association between noninvasive fibrosis markers and mortality among adults with nonalcoholic fatty liver disease in the United States. Hepatology. 2013;57:1357–1365. doi: 10.1002/hep.26156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Kim D, Konyn P, Sandhu KK, Dennis BB, Cheung AC, Ahmed A. Metabolic dysfunction-associated fatty liver disease is associated with increased all-cause mortality in the United States. J Hepatol. 2021;75:1284–1291. doi: 10.1016/j.jhep.2021.07.035. [DOI] [PubMed] [Google Scholar]
  • 10.Kim D, Vazquez-Montesino LM, Li AA, Cholankeril G, Ahmed A. inadequate physical activity and sedentary behavior are independent predictors of nonalcoholic fatty liver disease. Hepatology. 2020;72:1556–1568. doi: 10.1002/hep.31158. [DOI] [PubMed] [Google Scholar]
  • 11.Rinella ME, Lazarus JV, Ratziu V, Francque SM, Sanyal AJ, Kanwal F, et al. A multisociety Delphi consensus statement on new fatty liver disease nomenclature. J Hepatol. 2023;79:1542–1556. doi: 10.1016/j.jhep.2023.06.003. [DOI] [PubMed] [Google Scholar]
  • 12.Kim GA, Moon JH, Kim W. Critical appraisal of metabolic dysfunction-associated steatotic liver disease: implication of Janus-faced modernity. Clin Mol Hepatol. 2023;29:831–843. doi: 10.3350/cmh.2023.0277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Kim D, Ahmed A. Nonalcoholic fatty liver disease in early life and all-cause and cause-specific mortality. Hepatobiliary Surg Nutr. 2022;11:317–319. doi: 10.21037/hbsn-21-571. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Table 1.

Association between cardiovascular risk score and all-cause/cardiovascular mortality in individuals without MASLD (n=8,408)

Supplementary Table 2.

Association between cardiovascular risk score in MASLD and cancer-related mortality

Supplementary Table 3.

Assessment of cardiovascular risk score to predict all-cause and cardiovascular mortality based on the presence or absence of MASLD

SUPPLEMENTARY METHODS

Articles from Clinical and Molecular Hepatology are provided here courtesy of Korean Association for the Study of the Liver

RESOURCES