How much people drink is a heritable, continuously measured behavior — five studies between 2011 and 2019 trace the field from a first candidate-gene finding to a 480,842-person meta-analysis, distinct from the clinical diagnosis of alcohol use disorder.
What this condition connects to
Solid lines are connections this site curates. Dashed lines mean the two ends share a research paper — worth knowing, and not a claim that one explains the other.
Prevalence
Not applicable in the usual sense — alcohol consumption is a continuously measured behavioral trait, not a condition with a prevalence. The five studies behind this page range from 1,721 to 480,842 people, predominantly of European or, for one study, Korean ancestry.
Inheritance
Polygenic: the largest study behind this page found 46 new loci in 480,842 people, and no single variant — including AUTS2 and KLB, the two most mechanistically developed findings here — comes close to determining how much an individual drinks.
How much a person drinks, measured as grams of alcohol per day or similar, is a continuously distributed behavioral trait — related to, but distinct from, alcohol use disorder, the clinical diagnosis this site covers separately. This page follows five studies, spanning 2011 to 2019, that trace how genetic research on ordinary drinking behavior grew from a first small finding to population-biobank scale.
An early finding in Korean men, and the first major gene
Park et al. 2011 studied 1,721 Korean male drinkers, with replication in 1,113 more, and found a cluster of associated markers at 12q24. The same year, Schumann et al. 2011 ran a much larger European-ancestry meta-analysis — 26,316 people, replicated in 21,185 more — and found rs6943555 in AUTS2, a gene originally identified through autism research. The finding wasn't just statistical: the researchers found AUTS2 expression differed in human prefrontal cortex tissue and in the brains of mice specifically bred to drink different amounts of alcohol voluntarily — a real functional link, not only a population correlation.
A hormone from the liver, confirmed at biobank scale
Schumann et al. 2016 found rs11940694 in KLB, the gene for beta-Klotho — a protein that acts as an obligate co-receptor for FGF21, a hormone secreted by the liver already known to influence food preference. That gave alcohol-preference genetics a real mechanistic thread to follow, not just a locus. Clarke et al. 2017 then studied 112,117 UK Biobank participants and found 14 genome-wide-significant loci — independently confirming KLB at biobank scale with two of its own loci there, alongside alcohol-metabolizing genes ADH1B and ADH5, and CADM2, a gene that also turns up in this site's smoking- and grip-strength-related pages. Eight of this page's variants come from this study, the largest single source here.
Half a million people, and a neuropsychiatric connection
The largest effort, Schumann, Elliott et al. 2019, combined UK Biobank with two further consortia to reach 480,842 people of European descent — by far the largest sample behind any variant on this page. It found 46 new loci, followed up with brain MRI data and gene expression studies, and reported that some of the same genetic pathways behind ordinary drinking behavior are shared with neuropsychiatric disorders including schizophrenia — a finding the paper itself frames as a shared mechanism worth investigating further, not a claim that drinking causes or is caused by either condition.
Clinical detail
What this page does and does not describe
This page describes a continuously measured behavior, not a diagnosis, and none of these variants predict how much an individual drinks or whether they will develop a problem with alcohol. For the clinical diagnosis, see this site's separate alcohol use disorder page.
These are population-level statistical associations from self-reported drinking measures, each individually explaining only a small fraction of the variation between people. Environment, culture, availability, and individual circumstances shape drinking behavior at least as much as genetics does, and nothing here changes how alcohol use is clinically assessed.
Related variants MyGeneLog™ checks for
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about Alcohol Consumption comes down to these specific, well-studied positions — not a diagnosis. 85 positions are linked to this page; the ones this page's own text discusses are shown first.
The studies behind these variants recruited participants from different ancestries — a result found in one population doesn't always transfer to another. Based on 85 of 85 linked studies with a resolved discovery ancestry.
38.8%
58.8%
European · 38.8% East Asian · 2.4% Other named ancestries (NR, Asian unspecified, African American or Afro-Caribbean, Hispanic or Latin American, African unspecified) · 58.8%
No. This page covers how much people drink as a continuously measured behavior. This site's separate alcohol use disorder page covers the clinical diagnosis — a related but genuinely different question.
What is the AUTS2 finding about?
A 2011 study found a variant in AUTS2 — a gene first identified through autism research — associated with alcohol consumption, and confirmed it functionally: AUTS2 expression differed in human brain tissue and in mice bred to drink different amounts of alcohol voluntarily.
What is the KLB/FGF21 connection?
A 2016 study found a variant in KLB, the gene for beta-Klotho, a protein required for the liver hormone FGF21 to influence food and alcohol preference — a real mechanistic link later confirmed independently in a much larger 2017 UK Biobank study.
Does any variant here predict how much I drink?
No individual variant does. Even the largest study behind this page, with 480,842 people, found 46 loci that together explain only a small share of the variation in drinking behavior between people.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog™. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.