Table 2

Characteristics of included studies

Author /yearCountryStudy design and populationYrs f/uMethod of NAFLD diag*Method of measuring alcohol consumptionDefinition of moderate consumption studied as RFStudy outcomes of interest and event noAdjustments of interest consideredAdjusted HRs/OR/ mean differences for liver events with 95% CI and p valuesRisk of bias
Åberg 201956FinlandRetrospective data linkage cohort analysis
NAFLD population 6462, mean age 53 years, 60% M
10.9FLI>30Questionnaire at cohort entry<50 g/day in 10 g categories with abstinence as referenceComposite non-fatal and fatal liver disease
58 events
? unclear other than age, sexPer increase in 10 g of alcohol per day versus abstinence HR 1.43 (1.12 to 1.82) p=0.004High
Åberg et al 2020
57
Finland
(FINRISK Health survey)
Retrospective data linkage cohort analysis
NAFLD population 8345, mean age 53.7 years, 60% M
11.1FLI>60Questionnaire at cohort entry (recall for past month)<50 g/day in 10 g categories with abstinence as referenceComposite non-fatal and fatal liver disease 152 eventsAge, sex, smoking, T2DMg alcohol/day versus abstinence
0–9 hour 1.38 (0.74 to 2.58)
10–19 hour 2.18 (1.04 to 4.53)
20–29 hour 3.62 (1.67 to 7.76)
30–39 hour 3.53 (1.53 to 8.14)
40–49 HR 8.79 (3.95 to 19.56)
Low
Ajmera et al 2018
58
USARetrospective analysis of longitudinal cohorts within NASH CRN
NAFLD population 285, mean age 47 years, 30% M
3.9Liver biopsyQuestionnaire at cohort entry (Skinner lifetime drinking history)<2 drinks per day and excluded if >6 drinks on 1 occasion ≥monthlyHistological resolution or progressionon follow-up biopsyAge, sex, race, smokingPersistent moderate drinkers versus abstinence* resolution of NASH: OR 0.32 (0.11 to 0.92) p=0.04
fibrosis progression: adj mean diff 0.00 (−0.29 to 0.29) p=0.99
Mod
Chang et al 2019
59
South Korea
(Kangbuk Samsung Health Study)
Prospective population cohort
NAFLD population 58 927, mean age 37.7, 82% M
4.9USQuestionnaire at each study visit (annual or biennial)10–19.9 g/day (F)
10–29.9 g/day (M)
(low 1–9.9 g/day)
Fibrosis progress as estimated by high indirect serum scores**Age, sex, BMI, smoking, exercise level, education, T2DM, BPMod versus abstinence†
(repeat observations)
Fib4: HR 1.33 (1.13 to 1.57)
NFS: HR 1.37 (1.23 to 1.52) low versus abstinence
(repeat observations)
Fib 4: HR 1.08 (0.91 to 1.27)
NFS: HR 1.14 (1.02 to 1.27)
Low
Ekstedt et al 2009
60
SwedenRetrospective cohort
NAFLD population 71, mean age 47.3, 72% M
13.8US and liver biopsyQuestionnaire AUDIT-C and interview at follow-upg/day—no upper limit defined as ‘moderate’Fibrosis progress on follow-up biopsyAge, sex, BMI, T2DM, fibrosis at baselineIncreasing alcohol g/week versus abstinence
OR 1.012 (1.000 to 1.025) p=0.055
Low
Kawamura et al 201661JapanProspective cohort
NAFLD population 9959, mean age 49, 87% M
(included 18 patients >70 g alcohol/day defined as ARLD)
5.4USQuestionnaire at baseline and every 6 monthsg/day in categories with <20 g/day as referenceHCC on imagingAge, sex, BMI, T2DM, serum markersg/day alcohol versus <20 g/day
20–39 hour 0.90 (0.11 to 7.90) p=0.919
≥40–69 hour 2.48 (1.01 to 6.05) p=0.047
>70 hour 12.61 (5.68 to 28.00) p=0.001
Low
  • *Note multiple differences in means and OR presented for different histological and biochemical outcomes between abstainers, persistent moderate drinkers, and changes in alcohol consumption between biopsies. Presented data represent histological outcomes of potential clinical prognostic significance within the remit of this review comparing persistent moderate drinking to abstinence.

  • †Multiple HR presented in paper for different score outcomes for single and repeated outcome measures looking at intermediate/high or high-risk scores in low and moderate drinkers and different subgroups. Presented data represent outcomes best in keeping with remit of this review using widely used indirect serum markers of liver fibrosis.

  • ‡Scores used to estimate fibrosis progression were the Fib4 score, NAFLD fibrosis score (NFS) and AST to platelet ratio index (APRI) score.

  • ARLD, alcohol-related liver disease; AUDIT-C, alcohol use disorders identification test - consumption; BMI, body mass index; BP, blood pressure; CRN, clinical research network; FINRISK, Finland cardiovascular risk study; FLI, Fatty Liver Index; g, grams; HCC, hepatocellular carcinoma; M, Male; NAFLD, non-alcohol-related fatty liver disease; NASH, non-alcohol related steatohepatitis; RF, risk factor; T2DM, type 2 diabetes mellitus; US, hepatic ultrasound; Yrs, years.