Moderate Drinking Lowers Heart Risk but Raises Cancer Risk in 438,000 Veterans
A massive VA cohort finds J-shaped protection for heart disease but a linear cancer risk rise — even at moderate alcohol intake.
Summary
A prospective study of over 438,000 US Veterans found that moderate alcohol consumption (up to 30g/day) was associated with roughly 14–16% lower risk of death and major cardiovascular events compared to never-drinkers, following a J-shaped curve. However, cancer risk climbed linearly with every additional drink, with even moderate intake linked to elevated risk for alcohol-related cancers. Heavy drinking combined with smoking dramatically amplified head and neck cancer risk — more than five-fold versus non-drinking, non-smoking peers. The findings clarify the longstanding debate over alcohol's net health impact: cardiovascular benefits appear real at low doses, but they come at a meaningful oncological cost, meaning there is no truly 'safe' level of alcohol when cancer is factored into the equation.
Detailed Summary
Alcohol's relationship with health has been debated for decades, with cardiovascular benefit claims often used to justify moderate drinking. This large-scale cohort study from the Million Veteran Program provides some of the most statistically powerful evidence yet on alcohol's divergent effects across major disease categories.
Researchers analyzed self-reported alcohol consumption from 438,442 participants enrolled between 2011 and 2024 in the prospective MVP study. Outcomes tracked included total mortality, major adverse cardiovascular events (MACE — defined as myocardial infarction, stroke, or cardiovascular death), and incident cancers. Participants were categorized by daily alcohol intake from never-drinkers through heavy consumers exceeding 90g/day.
The cardiovascular and mortality results showed a classic J-shaped pattern. Current drinkers consuming 0.1–30g/day had a 14% lower risk of total mortality (HR 0.86) and a 16% lower risk of MACE (HR 0.84) compared to never-drinkers. This protective signal diminished and reversed at higher intakes. Cancer told a starkly different story: risk increased in a near-linear dose-response fashion, with heavy drinkers (>90g/day) facing a 28% higher cancer risk versus non-drinkers. Even the lowest drinking category showed a non-significant 3% elevation. The interaction between smoking and alcohol was particularly alarming for head and neck cancer, with current smokers who drank heavily facing a 5.69-fold elevated risk.
These findings suggest clinicians should not frame any level of alcohol as cardiovascular 'medicine.' While modest intake may confer some heart protection, the concurrent cancer liability — especially for head and neck, liver, and other alcohol-attributable cancers — means the net benefit calculation is far more nuanced than earlier studies implied.
Limitations include the observational design, reliance on self-reported intake, the predominantly male veteran population limiting generalizability, and the summary being based on the abstract only.
Key Findings
- Moderate alcohol (up to 30g/day) linked to 14% lower total mortality and 16% lower MACE risk vs. never-drinkers.
- Cancer risk rises linearly with alcohol dose; heavy drinkers (>90g/day) face 28% higher cancer risk.
- Smoking plus heavy drinking raises head and neck cancer risk 5.7-fold versus non-drinking, non-smoking peers.
- J-shaped curve for heart protection disappears at higher intakes, while cancer risk keeps climbing with no safe threshold.
- Even moderate drinking showed a non-significant 3% cancer risk increase, suggesting no fully safe lower limit.
Methodology
Prospective cohort study of 438,442 US Veterans from the Million Veteran Program (2011–2024), using self-reported alcohol consumption and tracking mortality, MACE, and incident cancer outcomes. Dose-response relationships were assessed using tests for both linear trend and curvature, with sex-stratified thresholds applied for moderate drinking definitions.
Study Limitations
This summary is based on the abstract only, as the full text was not accessible. The study population is predominantly male US Veterans, limiting generalizability to women and the general public. Observational design with self-reported alcohol intake introduces potential for misclassification and residual confounding.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: