Comparative associations of MASLD and MAFLD with the presence and severity of coronary artery calcification

We aimed to compare the associations of metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic dysfunction-associated fatty liver disease (MAFLD) with coronary artery calcification (CAC). Patients who simultaneously underwent ultrasonography to diagnose hepatic steatosis and cardiac computed tomography to detect CAC were included. The presence and severity of CAC were defined with CAC-score thresholds of >0 and > 300, respectively, and patients were divided into the following groups: no MASLD or MAFLD (reference), MASLD-only, MAFLD-only, and overlapping groups. Overall, 1,060/2,773 (38.2%) patients had CAC, of which 196 (18.5%) had severe CAC. The MASLD and MAFLD prevalence rates were 32.6% and 45.2%, respectively, with an overlap of 30.7%. In an ASCVD risk score-adjusted model, both MASLD (adjusted odd ratios [aOR], 1.21; 95% confidence interval [CI], 1.02–1.44; p = 0.033) and MAFLD (aOR, 1.20; 95% CI, 1.01–1.42, p = 0.034) were associated with CAC, whereas only MASLD (aOR, 1.38; 95% CI, 1.01–1.89, p = 0.041) was associated with severe CAC. Compared to the reference group, the overlapping group showed an association with CAC (aOR, 1.22; 95% CI, 1.01–1.47; p = 0.038); however, the MASLD and MAFLD subgroups did not differ in their association with CAC. MASLD may predict a higher risk of ASCVD more effectively than MAFLD.


Introduction
Non-alcoholic fatty liver disease (NAFLD) is de ned as hepatic steatosis of ≥ 5% without the evidence of excessive alcohol consumption and other chronic liver diseases 1,2 .It has a global prevalence of approximately 30-35% and coexists with obesity, insulin resistance, and diabetes, which accounts for a large socioeconomic burden 1,3 .The prognosis of NAFLD varies depending on the degree of brosis, with cardiovascular outcomes being the primary cause of mortality 1,4 .However, due to the heterogenous pathogenesis of NAFLD and uncertainty regarding its classi cation, NAFLD nomenclature underestimates the signi cance of metabolic comorbidities and extrahepatic manifestations 5 .
The nomenclature for metabolic dysfunction-associated fatty liver disease (MAFLD) was introduced in 2019 6 .Compared with NAFLD, MAFLD is a positive diagnosis that includes other chronic liver diseases and is categorised into three strati ed subtypes based on diabetes, obesity, and other metabolic components.Several studies have indicated that the MAFLD criteria may identify a higher risk of cardiovascular disease (CVD) than NAFLD [7][8][9] .However, different subtypes of MAFLD-such as diabetic and non-diabetic MAFLD-have different CVD morbidity and mortality rates 10,11 .Further drawbacks of the MAFLD nomenclature include heterogeneous pathogeneses resulting from mixed aetiologies and a restricted comprehension of the disease's natural history due to more lenient alcohol consumption 12,13 .
In 2023, metabolic dysfunction-associated steatotic liver disease (MASLD) emerged as the new nomenclature for steatotic liver disease (SLD), replacing both types of fatty liver disease, which have potentially stigmatising descriptions 12,13 .MASLD is de ned as the presence of at least one of ve cardiometabolic criteria of hepatic steatosis without any other apparent causes 12,13 .This umbrella term provides a comprehensive de nition of SLD, with a focus on metabolic dysfunction 12,13 .Due to the recent change in the nomenclature, the relationship between MASLD and the risk of CVD remains uncertain.Given the high predictive value of MAFLD for predicting the risk of CVD, it is important to elucidate the association between MASLD and CVD risk and compare it to that of MAFLD.
Quantitatively scoring coronary artery calci cation (CAC) is a reliable, non-invasive tool for detecting the presence of coronary artery disease and assessing the severity of CVD in asymptomatic individuals.This score is closely correlated with the degree of atherosclerosis and cardiovascular mortality, regardless of traditional risk factors 14,15 .
This study aimed to investigate the potential of MASLD and MAFLD as predictors of CVD risk based on the presence and severity of CAC.Additionally, we compared the associations between the MASLD and MAFLD subgroups with respect to the risk of CVD.

Baseline characteristics
Table 1 summarises the baseline characteristics of the patients with CAC (CAC group) and those without CAC (no-CAC group).Of the 2,773 participants, 1,060 (38.2%) were categorised into the CAC group.
Adjusted associations of CAC based on the presence or absence of MASLD or MAFLD Unlike the no-MAFLD group, the patients with MAFLD had an increased risk of severe CAC only after adjustment for age and sex (Model 1: aOR, 1.49; 95% CI, 1.10-2.03;p = 0.011), whereas no association with severe CAC was observed upon further adjustment (Table 3).

Adjusted associations of CAC according to subgroups of different steatotic liver statuses
We assessed the effect of CAC on four subgroups of patients categorised according to their steatotic liver status ( To directly compare the impact of CAC on each subgroup of patients with MASLD and/or MAFLD, we established the MASLD-only group as a reference (Fig. 2).Compared with the MASLD-only group, the MAFLD-only, MAFLD, and both MASLD and MAFLD groups were not signi cantly associated with CAC (all p > 0.05).

Discussion
In this study, we identi ed an association between MASLD and the presence and severity of CAC.In addition, we found that MASLD has a cardiovascular risk comparable to that of MAFLD but tends to outperform MAFLD in the assessment of CVD severity using the CAC score.This suggests that MASLD, as a novel nomenclature, has the potential to replace MAFLD when predicting the CVD risk.
This study has some clinical implications.First, the prevalence of MASLD in the two health-promotion centre populations was 32.6%.In two recent studies, the prevalence of MASLD was 33.4% in a Brazilian cohort and 14.87% in a Chinese cohort 16,17 .These discrepancies are possibly due to differences in the enrolled populations, including the timing of cohort recruitment and ethnic differences.Moreover, compared to a previous meta-analysis that reported the prevalence of MAFLD to be 38.77%(95% CI, 32.94-44.95%) 18, the prevalence of MAFLD in our study was slightly higher at 45.2%.Additionally, the prevalence of overlap between MASLD and MAFLD in our study was 30.7%, which was comparable to that between traditional NAFLD and MAFLD (27.4%) in a previous study 7 .Further research is required to investigate the prevalence of MASLD in a large population cohort that re ects sex-and ethnicity-speci c characteristics.
However, in the present study, MASLD showed an ASCVD risk score-adjusted association with severe CAC (OR, 1.38; 95% CI, 1.01-1.89),whereas MAFLD did not (OR, 1.15; 95% CI, 0.84-1.57).These results may be related to the heterogeneous subtypes of MAFLD.Among the three MAFLD subtypes, diabetes is the most well-established risk factor for CAC, and the MAFLD-diabetic subtype is associated with an increased risk of all-cause mortality and CVD 11,22 .However, recent research indicates that the MAFLDoverweight/obese subtype does not increase the risk of all-cause mortality or cardiovascular morbidity 10,11 .Additionally, the MAFLD-lean subtype with less than two metabolic factors is not associated with CAC 11 .To investigate the association between the presence of CAC and MAFLD subtypes, we divided the MAFLD group into diabetic and non-diabetic subgroups.We found that the presence of CAC was signi cantly associated with the MAFLD-diabetic subtype when adjusted for age and sex, but not with the MAFLD-non-diabetic subtype when unadjusted (Supplementary Table 1).This implies that the associations of overall MAFLD with severe CAC may differ due to its heterogeneous subtypes, indicating that MASLD may be more effective than MAFLD when evaluating the severity of CVD.Moreover, a previous study suggested that light and moderate alcohol consumption may have a protective effect on CVD mortality 23 .Therefore, MAFLD may be associated with a reduced risk of CAC due to the higher proportion of light-to-moderate drinkers than in patients with MASLD; however, further research is required to re ect the exact impact of alcohol intake on CAC.
Furthermore, our subgroup analysis revealed that only overlap between the MASLD and MAFLD groups was associated with the presence of CAC when neither group was used as a reference.However, the MASLD-only and MAFLD-only groups showed no association with CAC, which is a distinctive nding.Additionally, there was no evidence to suggest the superiority of CAC in the de nition of either type of SLD through comparisons between the MASLD-only and MAFLD subgroups.Contrary to our ndings, previous studies comparing the e cacies of NAFLD and MAFLD in predicting the risk of CVD showed that the MAFLD group and the overlap between the NAFLD and MAFLD groups were associated with an increased risk of CVD 7,9 .In one study, the aOR for a high ASCVD risk in the MAFLD-only and MAFLD groups were 3.26 and 3.14, respectively, using the NAFLD-only group as a reference group 24 .In addition, previous studies have shown that the MAFLD de nition is more effective than the NAFLD de nition when predicting the prognosis of other diseases 25,26 .
The subgroups in this study were characterised as follows: the MAFLD-only group included patients with metabolic risk factors and other aetiologies such as alcohol consumption and chronic hepatitis C, whereas the MASLD-only group had a comparatively favourable metabolic status, with all metabolic risk factors scoring 1.Compared to the de nition of NAFLD, the MASLD de nition is believed to more directly and comprehensively include the metabolic components of MAFLD, thereby homogenising the overlap of the metabolically unfavourable portions.In other words, groups other than the homogeneous groups (NAFLD-only and MAFLD-only) showed a favourable metabolic status when compared with the overlap group.Future studies should use large longitudinal data to determine the comparative associations between MASLD, MAFLD, and CVD risk.
In this study, we also compared the CAC risk of each MASLD and MAFLD adjusted for the presence of advanced brosis, de ned as Fib-4 ≥ 2.67, in addition to traditional risk factors.NAFLD is known to be associated with both fatal and nonfatal CVD events, depending on the stage of liver brosis 27 .We have previously shown that signi cant brosis, as de ned by MRE, is associated with the presence of CAC in patients with NAFLD 28 .Furthermore, considering that MAFLD may better identify signi cant liver brosis than NAFLD, the assessment of the brosis stage may be an important consideration when comparing CVD risk between MAFLD and MASLD 29 .
Our study had several limitations.First, this was a cross-sectional retrospective study that does not establish causality with CVD in patients with SLD.Second, we only included individuals who voluntarily visited the health-promotion centre and could afford the full test, including cardiac CT, which may have resulted in a selection bias.Third, this study was conducted in South Korea, potentially restricting the generalisability of the ndings to the entire population.Therefore, prospective, well-designed longitudinal studies are warranted to validate the causal relationship between MASLD and the incident and prevalent CVD risks.Fourth, the subjects enrolled in this study did not meet the criteria for MAFLD owing to the lack of homeostasis model assessment of insulin resistance (HOMA-IR) measurements.However, owing to the high cost and complexity of this procedure, the use of IR markers as the primary test for identifying MAFLD in the general population is challenging.In contrast, the MASLD can identify and assess risk factors through simple blood tests and demographic measurements.
In conclusion, beyond being associated with the presence of CAC, independent of traditional risk factors, MASLD was more effective than MAFLD in identifying severe CAC.Moreover, compared with the de nition of MAFLD, the new MASLD nomenclature might not show any gap in the prediction of CVD risk using the CAC score.Therefore, the assessment and strati cation of CAC may be useful in predicting a high risk of CVD events in patients with MASLD.

Patients
This cross-sectional, multicentre, retrospective study enrolled individuals who underwent comprehensive health screening using abdominal ultrasonography and cardiac computed tomography (CT) between January 2017 and December 2021 at two health-promotion centres in Daegu, South Korea.In total, 2,906 individuals who underwent abdominal ultrasonography and cardiac CT were examined.The exclusion criteria were a documented history of signi cant CVD or cardiac intervention and missing or inadequate data.Of the 2,906 patients, none experienced CVD events; however, 133 individuals had missing or inadequate data.Therefore, a total of 2,773 participants were enrolled in this study.

De nition and cation MASLD and MAFLD
Hepatic steatosis was assessed based on the following criteria: increased echogenicity of the liver compared with that of the renal cortex, deep attenuated beam, and blurred intrahepatic vessels on abdominal ultrasound 30 .MASLD is de ned as hepatic steatosis in addition to at least one of the ve cardiometabolic criteria 12,13 , whereas MAFLD is de ned as hepatic steatosis in addition to the presence of one of the following metabolic conditions: diabetes, overweight/obesity, or evidence of at least two out of seven metabolic abnormalities 6 .Due to all the enrolled patients being Korean, waist circumference and BMI were determined using cut-off values for Asians.HOMA-IR values was excluded due to the absence of fasting insulin levels.Individuals who did not ful l the MASLD or MAFLD criteria were classi ed into the no-MASLD or no-MAFLD groups, respectively.Moreover, individuals who met neither the MASLD nor the MAFLD criteria were classi ed into the neither MASLD nor MAFLD group.
Of the 2,773 participants, 905 met the criteria for MASLD, 1,253 met the criteria for MAFLD, 1,466 had neither MASLD nor MAFLD, and 851 met the criteria for overlap between MAFLD and MASLD.Moreover, 54 patients did not ful l the MAFLD criteria and were diagnosed with MASLD only, while 402 did not meet the MASLD criteria and were diagnosed with MAFLD only (Fig. 3).
Quanti cation of CAC using cardiac CT non-contrast cardiac prospective electrocardiogram-gated volumetric CT scans were performed at each institution using a 256-slice CT scanner (Revolution; GE Healthcare) and a 320-slice CT scanner (SOMATOM Force, Siemens Healthineers).At the end of inspiration, individuals were instructed to hold their breath while undergoing a scan ranging from the base of the heart to the carina.The tube voltage ranged from 100 to 120 kVp.Five lter revolutions were used for reconstruction, resulting in 2.5 mm-thick reconstruction slices.Subsequently, the CAC was calculated for each cardiac CT protocol.The CAC scores were calculated using the Agatston scoring method through independent postprocessing software (TeraRecon and Syngo.via;Siemens Healthineers).The presence of CAC was determined by a CAC score of > 0, whereas severe CAC was de ned as a score of > 300 14,31 .

Statistical analysis
Data are expressed as means with standard deviations, medians with interquartile ranges, numbers and percentages, as appropriate.Continuous data were analysed using the Student t-test or Mann-Whitney U test, and categorical data were compared using the chi-square test or Fisher exact test after testing for normality.To evaluate the associations of MASLD and MAFLD with the presence or severity of CAC, adjusted logistic regression models with stepwise backward elimination of OR were assessed.Model 1 was adjusted for age and sex, whereas Model 2 adjusted for smoking, statin use, diabetes, and advanced brosis, de ned as a brosis-4 index ≥ 2.67 after adjustment for Model 1.To avoid multicollinearity, Model 3 was adjusted for the ASCVD risk score only, which included age, sex, race, smoking status, blood pressure, use of antihypertensive medication, diabetes status, and levels of total and high-density lipoprotein cholesterol.
All statistical analyses were performed using the R software (version 4. The prevalence of MASLD and MAFLD.
Odds ratios for the presence of CAC between the subgroups of MASLD and MAFLD.
AbbreviationsALT alanine aminotransferase aOR adjusted odds ratio ASCVD atherosclerotic cardiovascular disease AST aspartate aminotransferase BMI body mass index CAC coronary artery calci cation CI con dence interval CT computed tomography CVD cardiovascular disease Fib-4 brosis-4 index HDL-C high-density lipoprotein cholesterol HOMA-IR Homeostasis model assessment of insulin resistance LDL low-density lipoprotein cholesterol MAFLD

Table 1
Baseline characteristics of the patients.Data are presented as mean ± standard deviation, median [interquartile range], or number (%).Presence of CAC is de ned as a

Table 2
shows the adjusted risk of CAC according to the presence or absence of MASLD or MAFLD.

Table 2
Adjusted risk for CAC according to the presence or absence of MASLD or MAFLD.*A p-value < 0.05.Model 1, adjusted for age and sex; Model 2, Model 1 + smoking, use of statins, diabetes, and advanced brosis; Model 3: † ASCVD risk score.Presence of CAC is de ned as a CAC score > 0 and advanced brosis is de ned as a brosis-4 index ≥ 2.67.† The ASCVD risk score includes age, sex, race, smoking status, blood pressure, medication use for hypertension, diabetes status, total cholesterol, and highdensity lipoprotein cholesterol.CAC, coronary artery calci cation; MASLD, metabolic dysfunctionassociated steatotic liver disease; MAFLD, metabolic dysfunction-associated fatty liver disease; aOR, adjusted odds ratio; CI, con dence interval; ref., reference; ASCVD, atherosclerotic cardiovascular disease.

Table 3
shows the adjusted associations of CAC severity in patients with MASLD and MAFLD compared to those without.Compared with the no-MASLD group, the patients with MASLD maintained a higher risk

Table 3
Adjusted risk of severe CAC according to the presence or absence of MASLD or MAFLD.*A p-value < 0.05.Model 1, adjusted for age and sex; Model 2, Model 1 + smoking, use of statins, diabetes, and advanced brosis; Model 3: † ASCVD risk score.Presence of severe CAC is de ned as a CAC score > 300 and advanced brosis is de ned as a brosis-4 index ≥ 2.67.† The ASCVD risk score includes age, sex, race, smoking status, blood pressure, medication use for hypertension, diabetes status, total cholesterol, and high-density lipoprotein cholesterol.CAC, coronary artery calci cation; MASLD, metabolic dysfunction-associated steatotic liver disease; MAFLD, metabolic dysfunction-associated fatty liver disease; aOR, adjusted odds ratio; CI, con dence interval; ref., reference; ASCVD, atherosclerotic cardiovascular disease.

Table 4
Adjusted risk of CAC according to the subgroups of different steatotic liver statuses.*A p-value < 0.05.Model 1, adjusted for age and sex; Model 2, Model 1 + smoking, use of statins, diabetes, and advanced brosis; Model 3: † ASCVD risk score.Presence of CAC is de ned as a CAC score > 0 and advanced brosis is de ned as a brosis-4 index ≥ 2.67.† The ASCVD risk score includes age, sex, race, smoking status, blood pressure, medication use for hypertension, diabetes status, total cholesterol, and highdensity lipoprotein cholesterol.CAC, coronary artery calci cation; aOR, adjusted odds ratio; CI, con dence interval; MASLD, metabolic dysfunction-associated steatotic liver disease; MAFLD, metabolic dysfunctionassociated fatty liver disease; ref., reference.