Background: Individuals with low socioeconomic status (SES) experience disproportionately greater alcohol-attributable health harm than individuals with high SES from similar or lower amounts of alcohol consumption. Our aim was to provide an update of the current evidence for the role of alcohol use and drinking patterns in socioeconomic inequalities in mortality, as well as the effect modification or interaction effects between SES and alcohol use, as two potential explanations of this so-called alcohol-harm paradox.

Methods: We did a systematic review, searching Embase, Medline, PsycINFO, and Web of Science (published between Jan 1, 2013, and June 30, 2019) for studies reporting alcohol consumption, SES, and mortality. Observational, quantitative studies of the general adult population (aged ≥15 years) with a longitudinal study design were included. Two outcome measures were extracted: first, the proportion of socioeconomic inequalities in mortality explained by alcohol use; and second, the effect modification or interaction between SES and alcohol use regarding mortality risks. This study is registered with PROSPERO (CRD42019140279).

Findings: Of 1941 records identified, ten met the inclusion criteria. The included studies contained more than 400 000 adults, more than 30 000 deaths from all causes, and more than 3000 100% alcohol-attributable events. Alcohol use explained up to 27% of the socioeconomic inequalities in mortality. The proportion of socioeconomic inequalities explained systematically differed by drinking pattern, with heavy episodic drinking having a potentially significant explanatory value. Although scarce, there was some evidence of effect modification or interaction between SES and alcohol use.

Interpretation: To reduce socioeconomic inequalities in mortality, addressing heavy episodic drinking in particular, rather than alcohol use in general, is worth exploring as a public health strategy.