Risk thresholds for alcohol consumption: combined IPD analysis of 599,912 current drinkers Strong AlcoholConfounding riskObservational evidence Observational study Large IPD analysis: mortality lowest around or below 100 g ethanol/week; higher intake progressively associated with lower life expectancy Causal status: Association; causal interpretation complicated by abstainer/former-drinker bias Reliability / bias notes: Former-drinker bias; drinking pattern and socioeconomic confounding Wood et al · Lancet · 2018 · PMID 29676281 Read the study →