Prevention & lipidsCohort study
Reassessing alcohol consumption and cardiovascular disease by addressing bias in observational data: results from the multi-ethnic study of atherosclerosis
- Prospective cohort study
- multi-ethnic
- ITT reallocation of former drinkers
- occasional drinkers as reference
Authors' conclusion
Associations between alcohol and CVD in observational studies are sensitive to methodologic variation.
Study context
- Population
- Adults aged 45–84 years without baseline cardiovascular disease (CVD), Multi-Ethnic Study of Atherosclerosis (MESA), 2000–20026,755 included
- Participants
- 6,814 adults
- Exposure
- Alcohol consumption categories, conventional vs bias-adjusted models
- Outcome
- CAC (0, 1-100, 101-300, >300) and clinical events (MI, stroke, CV mortality, MACE, HF)
Main finding
As reported in the abstract
CVD outcomes in bias-adjusted models
Each group vs occasionalHazard ratio (95% CI)
ModerateMI; attenuated, non-significant
0.64(0.38–1.06)
ModerateCV mortality; attenuated, non-significant
0.81(0.53–1.25)
LightMI
1.46(1.11–1.92)
Lightstroke
1.54(1.17–2.03)
LightMACE
1.33(1.12–1.57)
LightHF
1.42(1.10–1.84)
Observational data: associations, not effects.
Further findings
Drinking status distribution
Original publication
Reassessing alcohol consumption and cardiovascular disease by addressing bias in observational data: results from the multi-ethnic study of atherosclerosis. Eur J Prev Cardiol. Published 9 May 2026. doi:10.1093/eurjpc/zwag201
Summarised from the published abstract; every number is checked against it