Alcohol Related Liver Disease / Metabolic Dysfunction Associated Steatotic Liver Disease / Steatotic Liver Disease · Journal article

Steatotic liver disease subtypes and risk of hospitalization for sepsis: a nationwide cohort study.

Annals of Medicine · July 25, 2026

Strong Evidence Why this matters to you Expert discussion

Well-designed and adequately powered for the question it asks.

AI Summary

This is a large nationwide cohort study comparing sepsis hospitalization risk across three steatotic liver disease subtypes. All subtypes showed increased risk compared to non-SLD controls, with alcohol-related liver disease conferring the highest adjusted hazard ratio of 1.53, followed by metabolic dysfunction–alcohol-related disease (1.11) and metabolic dysfunction-associated disease (1.07), with associations persisting in comorbidity-free subgroups.

What was studied

Retrospective cohort study. Adults aged ≥20 years in South Korea who participated in the National Health Screening Program in 2012; stratified by presence and subtype of steatotic liver disease.. Intervention: Steatotic liver disease subtypes (MASLD, MetALD, ALD) defined by fatty liver index ≥30 and clinical characteristics. Compared with: Non-SLD reference group (fatty liver index <30). n = 4,389,734. South Korea (nationwide).

Key findings

Sepsis incidence per 1,000 person-years: ALD 4.49, MASLD 3.11, MetALD 2.25, non-SLD 2.20 Adjusted hazard ratios (aHR) for sepsis hospitalization: ALD 1.53 (p<0.001), MetALD 1.11 (p<0.001), MASLD 1.07 (p<0.001) In comorbidity-free sensitivity analysis: ALD aHR 1.60, MetALD 1.16, MASLD 1.09 (all p<0.001)

Safety

Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.

Who this may help

Clinicians managing patients with steatotic liver disease should recognize that all subtypes carry elevated sepsis risk, with alcohol-related disease carrying substantially higher risk than metabolic subtypes. This supports subtype-specific risk stratification and may inform infection prevention strategies and clinical surveillance protocols, particularly in ALD patients.

Plain-language summary

Large nationwide retrospective cohort with 4.4 million participants, 10-year follow-up, hard clinical endpoint (sepsis hospitalization), and multivariable adjustment showing distinct dose-response relationships across SLD subtypes with consistent results in sensitivity analyses.

Study design

As stated by the source record.

MethodsCohort of 4,389,734 adults who underwent health screening in 2012 enrolled from the Korean National Health Insurance Service database. SLD defined as fatty liver index ≥30 and classified into MASLD, MetALD, or ALD subtypes based on cardiometabolic risk profiles and alcohol intake; FLI <30 served as non-SLD reference. Sepsis identified via hospitalization claims and ICD-10 codes from 2013 to 2022. Adjusted hazard ratios estimated using multivariable Cox proportional hazards models adjusting for established sepsis risk factors, with subgroup and sensitivity analyses performed.

DesignRetrospective cohort study

PopulationAdults aged ≥20 years in South Korea who participated in the National Health Screening Program in 2012; stratified by presence and subtype of steatotic liver disease.

Sample size4,389,734

InterventionSteatotic liver disease subtypes (MASLD, MetALD, ALD) defined by fatty liver index ≥30 and clinical characteristics

ComparatorNon-SLD reference group (fatty liver index <30)

SettingSouth Korea (nationwide)

Reported figures

Quoted from the source exactly as published.

MASLD sepsis incidence3.11 per 1,000 person-years

MetALD sepsis incidence2.25 per 1,000 person-years

ALD sepsis incidence4.49 per 1,000 person-years

Non-SLD sepsis incidence2.20 per 1,000 person-years

Adjusted hazard ratio, ALD1.53

Adjusted hazard ratio, MetALD1.11

Adjusted hazard ratio, MASLD1.07

p-value, all subtypes<0.001

Clinical bottom line

Caveats

SLD definition relies on fatty liver index (FLI ≥30) rather than imaging or biopsy; validation of this classification method not reported.
Sepsis coding based on hospitalization claims and ICD-10 codes; accuracy of sepsis identification and severity stratification not validated.
No confidence intervals reported for adjusted hazard ratios.
Study population is South Korean; generalizability to other populations and health systems unclear.

Evidence assessment

Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.

Study design

strong

Nationwide retrospective cohort with 10-year prospective follow-up from administrative database.

Directness

strong

Hard clinical endpoint of sepsis hospitalization using ICD-10 codes; directly addresses clinical question.

Consistency

strong

Sensitivity analyses restricted to comorbidity-free individuals showed persistent associations with same direction and magnitude.

Risk of bias

moderate

Administrative database with complete follow-up; no mention of validation of SLD classification or sepsis coding.

Sample adequacy

strong

4.4 million participants with 79,445 to 1,332,944 per SLD subtype; sufficient for stratified analysis.

Source abstract

Background. The recent reclassification of steatotic liver disease (SLD) into metabolic dysfunction-associated SLD (MASLD), MASLD with increased alcohol intake (MetALD), and alcohol-related liver disease (ALD) provides a clinically relevant framework for distinguishing heterogeneous etiologic subtypes. However, whether these newly defined SLD subtypes differ in their susceptibility to sepsis remains unclear. We assessed the incidence and risk of hospitalization for sepsis across the SLD spectrum in a nationwide cohort.Methods. We conducted a retrospective cohort study using the Korean National Health Insurance Service database, including 4,389,734 adults who underwent health screening in 2012. SLD was defined as a fatty liver index (FLI) ≥30 and subtyped as MASLD, MetALD, or ALD based on cardiometabolic risk profiles and alcohol intake; FLI <30 defined the non-SLD reference group. Sepsis was identified using hospitalization claims and ICD-10 codes from 2013 to 2022. Adjusted hazard ratios (aHRs) were estimated using multivariable Cox proportional hazards models, with subgroup and sensitivity analyses.Results. The cohort comprised MASLD (n = 1,332,944; 30.4%), MetALD (n = 164,387; 3.7%), and ALD (n = 79,445; 1.8%). Sepsis incidence per 1,000 person-years was 3.11, 2.25, and 4.49, respectively, compared with 2.20 in the non-SLD group. After adjusting for established sepsis risk factors, all subtypes remained associated with higher risk, greatest in ALD (aHR 1.53), followed by MetALD (1.11) and MASLD (1.07) (all p < 0.001). MetALD showed lower crude incidence than MASLD but a higher adjusted risk in multivariable models. In sensitivity analyses restricted to individuals without comorbidities, associations persisted: ALD (aHR 1.60), MetALD (1.16), and MASLD (1.09) (all p < 0.001).Conclusions. All SLD subtypes were associated with increased but distinct risks of hospitalization for sepsis, with ALD conferring the highest risk. These findings support subtype-specific risk stratification and may inform infection-prevention strategies and clinical surveillance in individuals with SLD.

Follow from this

Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.