Document resource
WHAT IS ALREADY KNOWN ON THIS TOPIC The relationship between hepatic fibrosis and adverse atherosclerotic cardiovascular disease (ASCVD) risk in type 2 diabetes mellitus (T2DM) patients remains less known, particularly regarding the role of metabolic factors in modifying this association.WHAT THIS STUDY ADDS The hepatic fibrosis assessed by Steatosis-Associated Fibrosis Estimator (SAFE) score showed a significant and independent association with elevated 10-year ASCVD risk in T2DM patients, with this relationship moderated by hypertension, insulin resistance and low-density lipoprotein (LDL) cholesterol levels.HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY The cost-effective, non-invasive SAFE score may be used to identify T2DM patients who are at high risk of atherosclerotic cardiovascular disease. When integrated with metabolic parameters (including hypertension, insulin resistance indices, and LDL cholesterol), the combined approach improves risk stratification for patients with coexisting metabolic dysfunction-associated steatotic liver disease and T2DM.Introduction Metabolic dysfunction-associated steatotic liver disease (MASLD) affects approximately 25%–30% of adults globally and has become a major public health concern due to its escalating prevalence and adverse cardiometabolic consequences. 1 2 Notably, cardiovascular disease (CVD) accounts for over 40% of mortality in patients with MASLD, surpassing liver-related complications.3 4 Central to this association is hepatic fibrosis, a histopathological hallmark of MASLD progression, which has recently been recognized as an independent predictor of adverse cardiovascular outcomes (eg, myocardial infarction, heart failure).5 6Liver biopsy remains the gold standard for diagnosing hepatic fibrosis in MASLD, but its clinical utility is limited by its invasive nature and associated procedural risks. Therefore, non-invasive tools such as the Steatosis-Associated Fibrosis Estimator (SAFE) score have gained attention for their ability to assess fibrosis severity while incorporating metabolic parameters.7 Compared with traditional metrics like the fibrosis-4 index (FIB-4), the SAFE score is superior because it integrates metabolic risk factors (eg, body mass index (BMI), diabetes status) with liver biomarkers, thereby capturing both hepatic fibrosis severity and systemic metabolic dysregulation.7 While FIB-4 strongly correlates with CVD risk,8 9 the SAFE score’s inclusion of key metabolic factors suggests it may provide a more comprehensive measure of the interplay between hepatic fibrosis and CVD risk. However, no studies have directly evaluated the associations between SAFE score and CVD risk, leaving a critical gap in the literature.Understanding the relationship between hepatic fibrosis and CVD risk is particularly crucial in patients with type 2 diabetes mellitus (T2DM), a population with a high prevalence of both MASLD (up to ~70%) and CVD (a 2–4-fold increased risk compared with the general population).10–12 Mechanistically, the interplay between hepatic fibrosis and CVD in T2DM may be amplified by synergistic metabolic disturbances, including insulin resistance, dyslipidemia and chronic low-grade systemic inflammation.13–15 However, evidence directly linking hepatic fibrosis severity to CVD risk in patients with T2DM remains limited, especially regarding the role of metabolic factors in modifying this relationship.Quantitative risk assessment tools play a critical role in evidence-based primary CVD prevention by estimating an individual’s likelihood of experiencing a cardiovascular event. One of the most widely endorsed tools is the Atherosclerotic Cardiovascular Disease (ASCVD) Risk Calculator,16 which was developed by the American Heart Association (AHA) and American College of Cardiology (ACC). Building on this framework, our multisite cross-sectional study aimed to investigate whether hepatic fibrosis estimated by SAFE score is associated with an elevated 10-year ASCVD risk in a large cohort of patients with T2DM. Additionally, we sought to elucidate the potential moderating effects of key metabolic factors on the relationship between the hepatic fibrosis and ASCVD risk.Research design and methods Study population and design We conducted a cross-sectional, multisite study involving patients from three campuses of Sir Run Run Shaw Hospital, College of Medicine, Zhejiang University. The study population consisted of patients with T2DM and comorbid hepatic steatosis who were hospitalized in the endocrinology department between 2022 and 2024. Data were extracted from the electronic medical record systems of the participating centers.Inclusion criteria for this study were: (1) adults aged 40–79 years with a diagnosis of T2DM based on the American Diabetes Association criteria17; (2) presence of hepatic steatosis determined by ultrasonography. Exclusion criteria were (1) presence of viral hepatitis or hepatic cirrhosis of any etiology; (2) alcohol consumption ≥210 g/week for men or ≥140 g/week for women18; (3) pregnancy; (4) diagnosis of hyperthyroidism, hematologic malignancies, active cancer, pre-existing ischemic heart disease, stroke or history of coronary or peripheral revascularization procedures. A total of 1238 patients were included in the study.Demographic and clinical data Demographic variables included age and sex. Anthropometric measures included BMI and waist circumference (WC). The duration of diabetes history and blood pressure were recorded. Use of the following medications was also recorded: glucagon-like peptide-1 receptor agonists (GLP-1RAs), antihypertensive, antiplatelet and lipid-lowering medications. In addition, we recorded the number and type of antidiabetic drugs (OADs), including sulfonylureas, meglitinides, metformin, thiazolidinediones (TZDs), alpha-glucosidase inhibitors, dipeptidyl peptidase-4 inhibitors and sodium-glucose cotransporter-2 inhibitors (SGLT2i). Lifestyle factors included smoking status (never smoker, former smoker or current smoker) and alcohol consumption (categorized as none or mild drinker (<210 g/week for men or <140 g/week for women)).Laboratory measures Fasting blood samples were collected on the second day after hospital admission to measure the following: lipid profile, fasting blood glucose, hemoglobin A1c (HbA1c), globulin, albumin, platelet, uric acid, thyroid-stimulating hormone (TSH), high-sensitivity C reactive protein (hs-CRP), serum creatinine and liver enzymes (aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma-glutamyl transferase (GGT)). The Chronic Kidney Disease Epidemiology Collaboration equation was utilized to calculate the estimated glomerular filtration rate (eGFR). 19Insulin resistance estimation The estimated glucose disposal rate (eGDR, mg/kg/min) was used as a measure of insulin resistance, calculated using the following formula: eGDR (mg/kg/min)=21.158−(0.09×WC)−(3.407×hypertension)−(0.551×HbA1 c) (WC (cm), hypertension (yes=1/no=0) and HbA1c (%)). 20Diagnosis of hepatic steatosis, MASLD and fibrosis Hepatic steatosis was diagnosed using abdominal ultrasonography, with the following characteristic sonographic features: (1) marked hepatorenal echogenicity contrast, (2) diffuse hyperechogenicity of the liver parenchyma, (3) poorly defined intrahepatic vascular borders and/or (4) attenuated diaphragmatic acoustic interfaces. 21MASLD was defined as the presence of hepatic steatosis combined with at least one of five cardiometabolic risk factors, in the absence of other causes of hepatic steatosis. Cardiometabolic risk factors included: (1) BMI ≥23 kg/m² (for Asians) or WC ≥94 cm (for men) or 80 cm (for women); (2) fasting serum glucose ≥5.6 mmol/L, or 2-hour postload glucose ≥7.8 mmol/L, or HbA1c ≥5.7% or T2DM or ongoing T2DM treatment; (3) blood pressure ≥130/85 mm Hg or specific antihypertensive drug treatment; (4) plasma triglycerides ≥1.70 mmol/L or lipid-lowering treatment; (5) plasma HDL cholesterol ≤1.0 mmol/L (men) or 1.3 mmol/L (women) or lipid-lowering treatment.1Because all recruited patients had T2DM with hepatic steatosis, they met MASLD criteria after other causes of hepatic steatosis were excluded.The presence of advanced hepatic fibrosis was determined using SAFE score, calculated based on age (years), BMI (kg/m²), diabetes status (yes/no), AST (U/L), ALT (U/L), globulin (g/dL) and platelet count (×10⁹/L).7 In the current study, patients with SAFE score >100 were defined as having advanced fibrosis.7 22Estimation of the 10-year risk for developing ASCVD The ASCVD risk calculator was used to estimate the 10-year risk of developing a first fatal or nonfatal ASCVD event. Based on the 2013 ACC/AHA Pooled Cohort Equations, the 10-year ASCVD risk was calculated for each participant using the following nine variables: age, sex, race, systolic blood pressure, total cholesterol, high-density lipoprotein (HDL), diabetes history, smoking status and hypertension treatment. 16 We also categorized participants into ASCVD risk groups: low/borderline risk (≤7.4%), intermediate risk (7.5%–19.9%) and high risk (≥20%).16Statistical analysis We performed independent-samples t-test (for continuous variables) and χ 2 (for categorical variables) to compare the characteristics of patients with and without advanced hepatic fibrosis. We also performed one-way analysis of variance with Tukey HSD post-hoc tests (for continuous variables) and χ2 (for categorical variables) to compare the characteristics of patients with different levels of 10-year ASCVD risks.We first considered three potential indicators of fibrosis, namely SAFE score, FIB-423 and NFS.24 All three indicators were significantly associated with 10-year ASCVD risk after adjusting for BMI, sex, diabetes duration (years), HbA1c, smoking status (current/former vs never), alcohol consumption (mild drinker vs never), number of OADs, use of OADs (SGLT2i, GLP-1RAs or TZDs), use of statins and presence of kidney disease (eGFR <60 mL/min/1.73 m²). Among them, SAFE score demonstrated the strongest correlation with 10-year ASCVD risk (rp=0.461, p<0.001), outperforming FIB-4 (rp=0.383, p<0.001) and NFS (rp=0.425, p<0.001). Given its superior predictive value, SAFE score was selected for subsequent analyses, as it represented a better proxy for the systemic effects of hepatic fibrosis on cardiovascular health, enhanced statistical power, and increased the likelihood of detecting meaningful associations.We conducted linear regression models to examine the association between SAFE score and 10-year ASCVD risk. Next, to investigate whether metabolic factors moderated the relationship between SAFE score and 10-year ASCVD risk, we included each potential moderator and its interaction with SAFE score in the regression model. For metabolic factors that showed significant interactions, we further conducted stratified regression analyses within each metabolic group (e.g., hypertension group vs no hypertension group) to better elucidate the nature of the interaction.We evaluated ten metabolic factors as potential moderators: overweight/abdominal obesity, hypertension, triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), hs-CRP, insulin resistance (eGDR), uric acid, TSH, and HbA1c. Each metabolic factor was dichotomized based on established guidelines or median split. Specifically, overweight/abdominal obesity was defined as either BMI≥23 kg/m² or WC≥90 cm for men and 80 cm for women.25 Hypertension was defined as having a diagnosis of hypertension, use of anti-hypertensive medications, or blood pressure≥130/80 mmHg.26 Hypertriglyceridemia was defined as TG level≥150 mg/dL (1.70 mmol/L). Low HDL cholesterol was defined as HDL level<40 mg/dL (1.0 mmol/L) for men and<50 mg/dL (1.3 mmol/L) for women. High LDL level was defined as LDL level≥2.6 mmol/L. High hs-CRP was defined as hs-CRP level>2 mg/L. Insulin resistance was defined as eGDR≤5.56 mg/kg/min (median split), consistent with prior literature.27 Hyperuricemia was determined as serum uric acid level>7.0 mg/dL (420 µmol/L) for men and>6.0 mg/dL (360 µmol/L) for women.28 Elevated HbA1c was defined as≥7% for patients under 60 years old, ≥7.5% for patients between 60–69 years old, and ≥8.5% for patients aged 70 or above.29 30 Abnormal TSH was defined as TSH level<0.35 or >4.94 mIU/L.Covariates included sex, BMI, diabetes duration, HbA1c level, smoking status (current/former vs never), alcohol consumption (mild drinker vs never), number of OADs, use of OADs (SGLT2i, GLP-1RAs, or TZDs), use of statins, and presence of kidney disease (eGFR <60 mL/min/1.73 m²). Because age is a component of both the SAFE and the ASCVD score formula, and AST/ALT are components of the SAFE score, our multivariable models did not further adjust for age or AST/ALT to avoid overadjustment and multicollinearity. We retained BMI as a covariate given its strong confounding relationship with ASCVD risk in diabetes and consistency with prior literature.31 BMI was not included as a covariate for the analyses on overweight/abdominal obesity, and HbA1c was not included as a covariate for the analyses involving HbA1c. All analyses were performed in JMP Pro 16 and figures were created using GraphPad Prism V.10. The overall rate of missing data is low (1.3%) and we employed listwise deletion for missing data in analyses. All tests were two tailed with significance level of 0.05.Results Participant characteristics A total of 1238 hospitalized patients with T2DM and comorbid hepatic steatosis due to MASLD were included (37% female; mean age=57.81, SD=10.23 years; mean diabetes duration=9.1, SD=8.15 years; mean BMI=25.66, SD=3.58 kg/m²). Among all patients, 317 patients (25.6%) exhibited advanced fibrosis, while 921 individuals (74.4%) had no significant fibrosis.Table 1 presents the characteristics of patients categorized based on fibrosis severity. Patients with advanced fibrosis were older (64.59±9.51 vs 55.48±9.4 years, p<0.001), more likely to be female (45% vs 35%, p=0.002), had a higher BMI (26.92±4.15 vs 25.23±3.25 kg/m2, p<0.001) and a longer duration of diabetes (10.15±9.47 vs 8.75±7.61 years, p=0.018) compared with those without advanced fibrosis. Additionally, patients with advanced fibrosis exhibited higher metabolic risk markers, including higher systolic blood pressure, liver enzymes (AST, ALT, GGT), uric acid (p<0.001 for all).Table 1Clinical and biochemical characteristics, stratified by categories of hepatic steatosis with or without advanced fibrosisAll patients (N=1238)Patients without advanced fibrosis (SAFE≤100; n=921)Patients with advanced fibrosis (SAFE>100; n=317)Comparison between groups (P value)Age (years)57.81 (10.23)55.48 (9.40)64.59 (9.51)<0.001***Sex (female)463 (37%)321 (35%)142 (45%)0.002**Diabetes duration (years)9.10 (8.15)8.75 (7.61)10.15 (9.47)0.018*BMI (kg/m2)25.66 (3.58)25.23 (3.25)26.92 (4.15)<0.001***Waist circumference (cm)93.79 (9.09)92.82 (8.65)96.60 (9.74)<0.001***HbA1c (%)9.51 (2.21)9.50 (2.24)9.55 (2.12)0.722Fasting blood glucose (mmol/L)8.22 (2.62)8.20 (2.65)8.28 (2.56)0.630Current/former smoker433 (35%)335 (36%)98 (31%)0.077Mild alcohol consumer386 (31%)302 (33%)84 (26%)0.035*Hypertension666 (54%)444 (48%)222 (70%)<0.001***Systolic blood pressure (mm Hg)126.53 (14.10)125.67 (14.12)129.04 (13.75)<0.001***Diastolic blood pressure (mm Hg)78.42 (9.39)79.00 (9.51)76.74 (8.86)<0.001***AST (U/L)26.26 (18.60)21.77 (8.12)39.32 (30.54)<0.001***ALT (U/L)30.28 (25.04)26.27 (17.96)41.92 (36.50)<0.001***GGT (U/L)41.88 (45.28)37.03 (36.04)55.95 (63.06)<0.001***Globulin (g/L)26.52 (4.23)25.67 (3.86)28.98 (4.31)<0.001***Albumin (g/L)40.95 (3.63)41.05 (3.59)40.66 (3.72)0.102Platelet count (×10⁹/L)209.43 (57.64)221.81 (54.69)173.47 (50.55)<0.001***Triglyceride (mmol/L)2.24 (1.56)2.28 (1.61)2.11 (1.42)0.079HDL cholesterol (mmol/L)1.13 (0.26)1.12 (0.25)1.14 (0.29)0.483LDL cholesterol (mmol/L)3.05 (0.86)3.10 (0.85)2.91 (0.86)0.001**Serum uric acid (μmol/L)356.28 (95.21)351.61 (93.42)369.89 (99.16)0.005**Thyroid stimulating hormone (mIU/L)1.79 (1.80)1.69 (1.51)2.08 (2.44)0.009**hs-CRP (mg/L)4.35 (11.95)4.12 (12.49)5.01 (10.21)0.207Serum creatinine (μmol/L)73.20 (31.46)69.66 (26.20)83.47 (41.64)<0.001***Presence of kidney disease (eGFR<60 mL/min/1.73 m²)110 (9%)45 (5%)65 (21%)<0.001***Number of oral antidiabetic drugs (OADs)1.41 (1.21)1.44 (1.22)1.34 (1.18)0.233Use of OADs (SGLT2i, GLP-1RAs or TZDs)327 (26%)249 (27%)78 (25%)0.395Use of statins278 (22%)205 (22%)73 (23%)0.777Use of antiplatelet drugs97 (8%)57 (6%)40 (13%)<0.001***Use of antihypertensive drugs622 (50%)413 (45%)209 (66%)<0.001***10-year ASCVD risk score (%)17.83 (13.42)15.06 (11.15)25.87 (15.98)<0.001***Data presented as mean (SD) or N(%).*p<0.05, **p<0.01, ***p<0.001.ALT, alanine aminotransferase; ASCVD, atherosclerotic cardiovascular disease; AST, aspartate aminotransferase; BMI, body mass index; GGT, gamma-glutamyl transferase; GLP-1RAs, glucagon like peptide-1 receptor agonists; HDL, high-density liproprotein; hs-CRP, high sensitivity C-reactive protein; LDL, low-density lipoprotein; OADs, oral antidiabetic drugs; SAFE, Steatosis-Associated Fibrosis Estimator; SGLT2i, sodium-glucose cotransporter-2 inhibitors; TZDs, thiazolidinediones.Table 2 presents the characteristics of patients categorized based on 10-year ASCVD risk level. Patients with higher risk were older, had longer diabetes duration, larger WC, higher prevalence of hypertension and chronic kidney disease compared with lower risk groups. Among those with high CVD risk, the proportions of patients receiving antihypertensive drugs, antiplatelet drugs or lipid-lowering drugs were also higher.Table 2Clinical and biochemical characteristics, stratified by categories of the 10-year ASCVD risk calculator categoriesPatients with low/borderline ASCVD risk (n=349)Patients with intermediate ASCVD risk (n=426)Patients with high ASCVD risk (n=463)Comparisons between groups (p-value)Age (years)49.62 (6.83)*†55.53 (7.89)*‡66.09 (7.99)†‡<0.001***Sex (female)209 (60%)*†137 (32%)*‡117 (25%)†‡<0.001***Diabetes duration (years)7.15 (6.56)8.43 (7.49)11.19 (9.28)†‡<0.001***BMI (kg/m2)25.75 (4.06)26.08 (3.89)‡25.20 (2.75)‡0.001**Waist circumference (cm)91.87 (9.41)*†94.50 (9.77)*94.57 (7.91)†<0.001***HbA1c (%)9.54 (2.22)9.62 (2.22)9.39 (2.20)0.295Fasting blood glucose (mmol/L)8.45 (2.64)†8.29 (2.68)7.97 (2.54)†0.025*Current/former smoker26 (7%)*†173 (41%)*‡234 (51%)†‡<0.001***Mild alcohol consumer63 (18%)*†154 (36%)*169 (37%)†<0.001***Hypertension122 (35%)*†215 (50%)*‡329 (71%)†‡<0.001***Systolic blood pressure (mm Hg)121.81 (13.83)*†125.82 (13.19)*‡130.74 (13.89)†‡<0.001***Diastolic blood pressure (mm Hg)79.14 (9.74)78.41 (9.02)77.89 (9.45)0.171AST (U/L)26.62 (21.77)26.81 (17.30)25.49 (17.14)0.527ALT (U/L)32.30 (28.37)†32.58 (26.21)‡26.64 (20.48)†‡<0.001***GGT (U/L)39.77 (45.34)41.00 (36.77)44.28 (51.84)0.329Globulin (g/L)26.48 (4.28)26.40 (4.28)26.66 (4.16)0.647Albumin (g/L)41.50 (3.71)†41.38 (3.46)‡40.14 (3.59)†‡<0.001***Platelet count (×10⁹/L)221.22 (58.40)*†208.59 (54.63)*201.31 (58.39)†<0.001***Triglyceride (mmol/L)2.22 (1.51)2.27 (1.59)2.21 (1.57)0.826HDL cholesterol (mmol/L)1.16 (0.25)†1.12 (0.26)1.11 (0.26)†0.011*LDL cholesterol (mmol/L)3.09 (0.86)3.09 (0.85)2.99 (0.86)0.120Serum uric acid (μmol/L)344.55 (95.07)†360.11 (92.82)361.49 (96.90)†0.027*Thyroid stimulating hormone (mIU/L)1.91 (1.87)1.69 (1.81)1.80 (1.73)0.249hs-CRP (mg/L)4.38 (11.47)4.12 (11.54)4.53 (12.67)0.877Serum creatinine (μmol/L)60.65 (16.08)*†71.84 (30.44)*‡83.91 (36.98)†‡<0.001***Presence of kidney disease (eGFR<60 mL/min/1.73 m²)9 (3%)22 (5%)‡79 (17%)†‡<0.001***Number of oral antidiabetic drugs (OADs)1.32 (1.15)1.40 (1.27)1.49 (1.20)0.124Use of OADs (SGLT2i, GLP-1RAs or TZDs)86 (25%)122 (29%)119 (26%)0.415Use of statins60 (17%)94 (22%)124 (27%)†0.005**Use of antiplatelet drug10 (3%)*†29 (7%)*‡58 (13%)†‡<0.001***Use of hypertensive medications103 (30%)*†208 (49%)*‡311 (67%)†‡<0.001***SAFE score16.22 (74.07)*†45.74 (81.93)*‡81.71 (85.77)†‡<0.001***Data presented as mean (SD) or N(%).*p<0.05, **p<0.01, ***p<0.001.*Significant difference between low/borderline and intermediate groups†Significant difference between low/borderline and high groups.‡Significant difference between lintermediate and high groups.ALT, alanine aminotransferase; ASCVD, atherosclerotic cardiovascular disease; AST, aspartate aminotransferase; BMI, body mass index; GGT, gamma-glutamyl transferase; GLP-1RAs, glucagon like peptide-1 receptor agonists; HDL, high-density liproprotein; hs-CRP, high sensitivity C-reactive protein; LDL, low-density lipoprotein; OADs, oral antidiabetic drugs; SAFE, Steatosis-Associated Fibrosis Estimator; SGLT2i, sodium-glucose cotransporter-2 inhibitors; TZDs, thiazolidinediones.Association between SAFE score and 10-year ASCVD risk We examined the relationship between SAFE score and 10-year CVD risk through multivariable regression modeling. After full adjustment for covariates, each unit increase in SAFE score was significantly associated with a 0.07-point elevation in 10-year ASCVD risk (unstandardized coefficient B=0.07, SE=0.004; p<0.001; full results in table 3).Table 3Fibrosis was associated with higher 10-year ASCVD riskAll participants (n=1238)coefficient B (SE), pIntercept24.29 (2.90), <0.001***BMI (kg/m2)−0.65 (0.09), <0.001***Sex (female)−5.46 (0.79), <0.001***HbA1c (%)0.34 (0.15), 0.025*Diabetes duration (years)0.32 (0.04), <0.001***Number of oral antidiabetic drugs (OADs)0.14 (0.33), 0.659Current/former smoker (yes)7.50 (0.80), <0.001*Mild alcohol consumer (yes)−2.21 (0.77), 0.004**Use of OADs (SGLT2i, GLP-1RAs or TZDs)−0.68 (0.82), 0.411Presence of kidney disease (eGFR<60 mL/min/1.73 m²)6.08 (1.11), <0.001***Use of statins0.32 (0.77), 0.672SAFE score0.07 (0.004), <0.001****p<0.05, **p<0.01, ***p<0.001.BMI, body mass index; GLP-1RAs, glucagon like peptide-1 receptor agonists; OADs, oral antidiabetic drugs; SAFE, Steatosis-Associated Fibrosis Estimator; SGLT2i, sodium-glucose cotransporter-2 inhibitors; TZDs, thiazolidinediones.Metabolic moderators of the fibrosis–ASCVD relationship Further analyses investigated whether metabolic factors moderated the relationship between SAFE score and 10-year ASCVD risk. Hypertension, insulin resistance (eGDR ≤5.56 mg/kg/min, based on median-split) and LDL cholesterol significantly moderated the association between SAFE score and 10-year ASCVD risk after adjusting for covariates ( ps <0.05 for all interaction terms).Hypertension significantly moderated the relationship between SAFE score and 10-year ASCVD risk after adjusting for covariates, B=0.02, SE=0.01, p=0.010 (full results in online supplemental table 1). Stratified analyses revealed that among patients with hypertension, each unit increase in SAFE score was associated with a 0.07-point increase in 10-year ASCVD risk, B=0.07, SE=0.004, p<0.001 (figure 1A). Among patients without hypertension, each unit increase in SAFE score was associated with a 0.04-point increase in 10-year ASCVD risk, B=0.04, SE=0.01, p<0.001 (figure 1B). These findings suggest that hepatic fibrosis was associated with greater increase in CVD risk in patients with hypertension, compared with those without hypertension.SP110.1136/bmjdrc-2025-005135.supp1Supplementary dataFigure 1The adjusted relationship between SAFE score and 10-year ASCVD risk score in patients with hypertension (A) and patients without hypertension (B); patients with insulin resistance, defined as eGDR≤5.56 mg/kg/min based on median-split (C) and patients without insulin resistance, defined as eGDR>5.56 mg/kg/min (D); patients with high LDL (E) and patients without high LDL (F). The solid lines and dotted lines represent lines of best fit and 95% CIs, respectively. ASCVD, atherosclerotic cardiovascular disease; eGDR, estimated glucose disposal rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SAFE, Steatosis-Associated Fibrosis Estimator.Insulin resistance (defined as eGDR ≤5.56 mg/kg/min, based on median-split) also significantly moderated the relationship between SAFE score and 10-year ASCVD risk after adjusting for covariates, B=0.01, SE=0.01, p=0.045 (full results in online supplemental table 2). Stratified analyses revealed that among patients with insulin resistance (eGDR ≤5.56 mg/kg/min), each unit increase in SAFE score was associated with a 0.07-point increase in 10-year ASCVD risk score, B=0.07, SE=0.01, p<0.001 (figure 1C). Among patients without insulin resistance (eGDR >5.56 mg/kg/min), each unit increase in SAFE score corresponded to a 0.05-point increase in 10-year ASCVD risk, B=0.05, SE=0.005, p<0.001 (figure 1D). The findings suggest that hepatic fibrosis was associated with greater increase in ASCVD risk in patients with insulin resistance, compared with those without insulin resistance. To avoid bias from dichotomization, we repeated analyses treating eGDR as a continuous variable. Results remained consistent, with eGDR significantly moderating the association between SAFE score and 10-year ASCVD risk after adjusting for covariates (B=−0.01, SE=0.001, p<0.001).Moreover, high LDL significantly moderated the relationship between SAFE score and 10-year ASCVD risk after adjusting for covariates, B=0.02, SE=0.01, p=0.034 (full results in online supplemental table 3). Stratified analyses revealed that among patients with high LDL, each unit increase in SAFE score was associated with a 0.08-point increase in 10-year ASCVD risk score, B=0.08, SE=0.005, p<0.001 (figure 1E). Among patients without high LDL, each unit increase in SAFE score corresponded to a 0.05-point increase in 10-year ASCVD score, B=0.06, SE=0.01, p<0.001 (figure 1F). The findings suggest that hepatic fibrosis was associated with greater increase in ASCVD risk in patients with high LDL, compared with those without high LDL.Additionally, we conducted stratified analyses comparing participants who have all three metabolic risk factors (ie, hypertension, insulin resistance and high LDL) with those who do not have any of the three metabolic risk factors (ie, no hypertension, no insulin resistance, no high LDL). Among those with all three metabolic risk factors, each unit increase in SAFE score corresponded to a 0.06-point increase in 10-year ASCVD score after adjusting for covariates, B=0.06, SE=0.01, p<0.001 (figure 2A). Among those without any of the three metabolic risk factors, each unit increase in SAFE score corresponded to a 0.03-point increase in 10-year ASCVD score, B=0.03, SE=0.01, p=0.006 (figure 2B).Figure 2The adjusted relationship between SAFE score and 10-year ASCVD risk score in patients with hypertension, insulin resistance and high LDL (A); and patients without hypertension, insulin resistance or high LDL (B). The solid lines and dotted lines represent lines of best fit and 95% CIs, respectively. ASCVD, atherosclerotic cardiovascular disease; LDL, low-density lipoprotein; SAFE, Steatosis-Associated Fibrosis Estimator.Other metabolic factors did not moderate the fibrosis–ASCVD relationship The association between SAFE score and 10-year CVD risk was not moderated by other metabolic factors. Regression models showed no significant moderating effects of overweight/abdominal obesity (B=0.01, SE=0.02, p=0.515), hs-CRP (B=0.002, SE=0.01, p=0.764), TG (B=−0.01, SE=0.01, p=0.118), HDL cholesterol (B=−0.013, SE=0.01, p=0.053), uric acid (B=0.01, SE=0.01, p=0.260), TSH (B=−0.02, SE=0.02, p=0.333) or HbA1c (B=−0.01, SE=0.01, p=0.213), after adjusting for the covariates. Full results are presented in online supplemental table 4.Discussion This multisite cross-sectional study provides novel insights into the complex interplay between hepatic fibrosis, metabolic parameters and ASCVD risk in hospitalized patients with T2DM and MASLD, who had no pre-existing CVD. Our findings suggest that higher SAFE score independently predicted an elevated 10-year ASCVD risk in this high-risk population. Furthermore, we identified hypertension, insulin resistance and LDL cholesterol as significant metabolic moderators of the association between hepatic fibrosis and ASCVD risk. These results advance our understanding of the intricate cardiometabolic interplay in MASLD and carry important implications for ASCVD risk stratification strategies.MASLD with advanced fibrosis has been strongly associated with a heightened risk of cardiovascular morbidity and mortality, including myocardial infarction, heart failure and all-cause mortality.5 32 Accumulating research suggests that advanced hepatic fibrosis, as quantified by FIB-4, NFS and the AST to platelet ratio index (APRI), represents an independent risk factor for cardiovascular mortality beyond traditional cardiometabolic risk factors in the general population.9 33 34 Notably, Choi et al demonstrated a significant and independent association between these fibrosis indices and CVD mortality.6However, existing evidence primarily comes from studies on the general populations, including both diabetic and non-diabetic individuals, with a paucity of research focused exclusively on T2DM patients. A previous observational study utilizing Fatty Liver Index and BARD scores to assess hepatic steatosis and advanced fibrosis reported an association between hepatic fibrosis and CVD risk in newly diagnosed T2DM.35 However, this study lacked HbA1c measurements, precluding an evaluation of glycemic control and relied solely on biochemical indices for diagnosing hepatic steatosis, which are less accurate than hepatic ultrasonography. In contrast, our study incorporated comprehensive metabolic parameters, including HbA1c, and used ultrasonography to diagnose hepatic steatosis. Building on recent work by Li et al, which demonstrated the SAFE score’s superior diagnostic accuracy for severe fibrosis and liver-related events compared with FIB-4, NFS and APRI,36 our study revealed that the SAFE score demonstrates a stronger association with ASCVD risk compared with FIB-4 or NFS in hospitalized patients with T2DM. Importantly, this association persisted after comprehensive adjustment for traditional cardiovascular risk factors, underscoring the independent prognostic value of hepatic fibrosis severity as captured by SAFE score. To our knowledge, this is the first large-scale study to examine the association between the SAFE score and ASCVD risk.T2DM is intrinsically associated with multiple metabolic abnormalities, including obesity, insulin resistance, dyslipidemia, hypertension and hyperuricemia. Metabolic factors can independently or synergistically contribute to the development of both hepatic fibrosis and CVD.15 To better understand these interactions, we examined the moderating effects of key metabolic parameters on the relationship between hepatic fibrosis and ASCVD risk. We found that this relationship was significantly stronger in patients with hypertension, insulin resistance and with elevated LDL levels, consistent with prior research.37 A meta-analysis identified hypertension as a critical determinant in the link between hepatic steatosis and CVD.38 Furthermore, insulin resistance and hyperlipidemia are known shared pathological mechanisms underlying MASLD and CVD progression.39 40 These mechanisms likely explain our observation of amplified associations between hepatic fibrosis and ASCVD risk in patients with insulin resistance and elevated LDL levels. Collectively, these findings suggest that hypertension, insulin resistance and hyper-LDL-cholesterolemia may function as critical effect modifiers in the hepatic fibrosis–ASCVD risk relationship. Prospective cohort studies are warranted to confirm these observations and unravel the temporal dynamics of these metabolic interactions.Conversely, traditional lipid parameters (triglycerides and HDL cholesterol) did not significantly moderate the SAFE score–ASCVD risk association. One potential explanation is the high prevalence of statin use among patients, which may have attenuated the variability in triglyceride and HDL cholesterol levels, thereby blunting their moderating effects on ASCVD risk. Additionally, overweight/abdominal obesity was not a significant moderator, likely due to the fact that all participants had T2DM, a key diagnostic criterion for MASLD. The high frequency of obesity in this cohort may have limited the discriminative power in identifying its moderating effect on the hepatic fibrosis-ASCVD link. However, these non-significant findings do not negate the clinical relevance of obesity, dyslipidemia, uric acid, TSH, HbA1c or hs-CRP in CVD risk. Instead, they suggest that the influence of these metabolic factors on hepatic fibrosis–ASCVD interactions may be indirect, masked by stronger metabolic drivers (eg, insulin resistance), or inadequately captured by current categorical definitions.Clinical implications The prevention of CVD has become a global health priority, underscoring the need for cost-effective non-invasive evaluation tools. Traditional CVD risk models do not account for hepatic fibrosis, despite its established association with cardiovascular events. Our findings highlight the SAFE score as a potentially valuable metric for ASCVD risk prediction, demonstrating a stronger correlation with ASCVD risk than traditional fibrosis indices. Given its non-invasive nature and superior predictive value, the SAFE score could serve as a clinically actionable tool for identifying high-risk individuals, allowing for earlier cardiovascular intervention strategies in patients with MASLD.Furthermore, our results suggest that ASCVD risk stratification in MASLD should incorporate key metabolic factors, particularly hypertension, insulin resistance, and LDL cholesterol. A multimodal risk assessment framework, integrating hepatic fibrosis severity with metabolic risk profiling, may enhance MASLD management and optimize preventive strategies against MASLD-associated cardiovascular morbidity.Pathological mechanisms linking hepatic fibrosis and CVD risk Hepatic fibrosis and CVD share some pathophysiological mechanisms, such as disruption of lipid metabolic homeostasis, dysregulation of glucose metabolism, accumulation of oxidative stress damage, chronic systemic inflammation, neuroendocrine regulatory abnormalities, prothrombotic state formation and gut microbiota dysbiosis. 15 Of particular significance, this hepatic fibrosis–CVD relationship may stem from the progressive deterioration of systemic/hepatic insulin resistance, aberrant secretion of atherogenic lipoproteins and systemic release of multiple proinflammatory cytokines, prothrombogenic factors and vasoactive mediators. Hepatic-derived mediators exacerbate vascular endothelial dysfunction, accelerate atherosclerotic plaque formation, intensify cardiomyocyte metabolic stress and disrupt the coagulation–fibrinolysis balance, ultimately drives the development and progression of CVD.3 41–44 Nevertheless, the precise mechanisms underlying the association between hepatic fibrosis and CVD remain incompletely understood, necessitating further mechanistic investigations.Study limitations and strengths This study has several limitations that should be acknowledged. First, the observational design precludes causal inference between SAFE score and ASCVD development. Second, while ultrasonography and non-invasive fibrosis assessment are clinically practical, their diagnostic accuracy remains inferior to liver biopsy histology. Third, our study only included hospitalized patients, potentially introducing selection bias towards individuals with more poorly controlled diabetes. This may limit the generalizability of our findings to community-dwelling populations. Future prospective cohort studies incorporating longitudinal designs and histopathological validation will be crucial in further elucidating the hepatic fibrosis–ASCVD relationship. In addition, although the SAFE score has demonstrated diagnostic utility in general populations, it has not been specifically validated in T2DM populations using liver histology. This limitation should be considered when interpreting our findings. Age was embedded within the formulas for SAFE and ASCVD scores and AST/ALT were embedded within the formula for SAFE. We did not include them as separate covariates to prevent overadjustment and multicollinearity. Thus, findings should be interpreted as the association between these composite measures rather than effects independent of age or liver enzymes.Despite these limitations, our investigation demonstrates methodological rigor through its multisite design with substantial sample size, comprehensive data integrity and systematic exclusion of patients with active malignancies, hepatic cirrhosis or pre-existing CVD events to mitigate potential confounding effects on outcome interpretation.Conclusion In hospitalized patients with T2DM, hepatic fibrosis severity, as assessed by the SAFE score, is independently associated with 10-year ASCVD risk. Moreover, this association was significantly amplified by hypertension, insulin resistance and LDL cholesterol levels. These findings underscore the potential clinical utility of the SAFE score in cardiovascular risk prediction and emphasize the need for multidimensional metabolic risk assessment in MASLD management.