Baseline characteristics

Table 1 summarizes the baseline characteristics of our study population (n = 2,031,521), categorized into five groups: 789,654 (38.9%) participants with no SLD; 1,008,394 (49.6%) with MASLD; 147,192 (7.2%) with MetALD; 46,369 (2.3%) with ALD with MD and 39,912 (2.0%) with MASLD with VH.

Participants with MetALD or ALD with MD tended to be younger, male, and more likely to be a current smoker. In addition, the MetALD, and ALD with MD groups had higher levels of FG, TG, AST, ALT, and γ-GTP and lower LDL-C levels. The MetALD, and ALD with MD groups had lower rates of CKD, CCI sore ≥ 5, OHA ≥ 3, insulin use, statin use, aspirin use, and oral anticoagulant (OAC) use.

In contrast, participants without SLD were more likely to be older, female, and have a lower BMI, waist circumference, and blood pressure than those in the SLD group.

Risk of HF in different SLD subtypes

Table 2 shows the risk of incident HF and CV mortality according to SLD status. The median follow-up was 7.1 (6.0–8.1) years for HF and 7.2 (6.1–8.1) years for CV death. There was a higher risk of developing HF, compared to those with no SLD, for individuals with MASLD (aHR, 1.11; 95% CI, 1.09–1.13), MetALD (aHR, 1.14; 95% CI, 1.10–1.17), ALD with MD (aHR, 1.32; 95% CI, 1.26–1.38) and MASLD with VH (aHR, 1.12; 95%CI, 1.10–1.17) in model 4.

Table 2 Risk of HF and CV mortality in different SLD subtypesCV mortality risk by SLD subtypes

Individuals with MASLD had a 27% increased risk of CV mortality compared to the no SLD group (aHR,1.27; 95% CI, 1.21–1.33) Similarly, the risk of CV mortality was increased in MetALD (aHR, 1.13), ALD with MD (aHR, 1.25), although to a lesser extent than the MASLD group. The MASLD with VH groups showed no statistically significant difference with no SLD group.

Risk of HF by alcohol consumption and FLI categories

Table 3 presents the risk of incident HF and CV mortality stratified by the FLI and alcohol consumption. Among individuals with FLI < 30, nondrinkers were included in the reference group. In the low-FLI group (FLI < 30), mild alcohol consumption (30 g/day in male and < 20 g/day in female) was associated with a lower risk of HF (aHR, 0.87; 95% CI, 0.85–0.89). Conversely, very heavy alcohol consumption (≥ 60 g/day for male and ≥ 50 g/day for female) was associated with a higher risk of HF (aHR, 1.16; 95% CI, 1.05–1.28). This J-shaped trend persisted in the moderate-FLI (FLI 30–60) and high-FLI (FLI ≥ 60) groups as displayed in Fig. 2. Notably, the risk of new-onset HF increased with higher FLI levels regardless of alcohol consumption.

Table 3 Risk of HF and CV mortality by alcohol consumption and FLI categoriesFig. 2

Risk of HF and CV mortality by alcohol consumption and FLI

Risk of CV mortality by alcohol consumption and FLI categories

Table 3 shows an association between increasing FLI levels and higher CV mortality regardless of alcohol consumption. When plotting the relationship between FLI and the risk of HF, CV mortality, it showed a linear association presented to Fig. 3. This trend was particularly pronounced in non-alcohol group, where the risk of CV mortality progressively increased with higher FLI (FLI 30–60: aHR, 1.21; 95% CI, 1.14–1.28; FLI ≥ 60: aHR, 1.80; 95% CI, 1.67–1.95).

Fig. 3

Linear relationship between high FLI and increased risk of HF and CV mortality

However, mild drinkers (30 g/day in male and < 20 g/day in female) tended to have a lower risk of CV mortality than non-alcohol group. Even after excluding participants with VH, the results remained unchanged (Supplementary Table 3).

Risk of all-cause and liver-related mortality by SLD subtypes

Regarding all-cause and liver-related mortality, all subtype of SLD groups showed increased risk compared to no SLD group. (Table 4) In addition, MetALD and ALD with MD group showed more increased risk compared to MASLD. Interestingly, viral hepatitis showed additional impact on all-cause and liver-related mortality.

Table 4 Risk of all-cause and liver-related mortality in different SLD subtypesFig. 4

Impact of adjusting variables on risk of HF and CV mortality

Risk of all-cause mortality and liver-related mortality by alcohol consumption and FLI categories

Table 5 presented that higher FLI associated with higher risk of all-cause and liver-related mortality. Increased alcohol consumption (≥ 60 g/day for male and ≥ 50 g/day for female ) showed increased risk of all-cause and liver-related mortality.

Table 5 Risk of all-cause mortality and liver-related mortality by alcohol consumption and FLI categoriesSubgroup analysis

Supplementary Table 4 shows the subgroup analyses of the risks of HF and CV mortality stratified by several variables. We found that SLD increased the risk of HF or CV mortality, independent of the indicated factors (P for interaction > 0.05). However, the effect of SLD on the risk of HF was influenced by factors such as age, sex, BMI, CKD, smoking status, use of HTN medication, HDL level, use of statins, regular exercise, DM duration, fasting glucose level and use of OHA. In addition, several factors, such as age, HDL-C level, regular exercise, statin and use of OHA or insulin affected the association between SLD and CV mortality.

Sensitivity analysis

Supplementary Table 5 exhibits the risk of HF and CV mortality according to SLD status and history of CV disease. Although participants with a history of CV disease showed a higher risk of incident HF and CV mortality compared with participants without a history of CV disease, our analysis revealed that the spectrum of SLD increased the risk of HF and CV mortality regardless of history of CV disease. (P for interaction > 0.05)

Furthermore, when we performed competing risk analysis, as non-CV death was considered a competing risk, the SLD group showed consistently increased risk of HF, CV mortality and liver-related mortality as presented in Supplementary Table 6. Also Supplementary Tables 7 and Fig. 4 showed the HRs for the adjusting variable.