Study overturns idea that light drinking protects brain health
A large-scale study combining observational and genetic data has found that any alcohol consumption increases dementia risk, with no safe level identified. Published in BMJ Evidence Based Medicine, the research challenges previous beliefs about moderate drinking's benefits. Dementia risk rises linearly with alcohol intake, according to the analysis.
Researchers analyzed data from 559,559 participants in the US Million Veteran Program and UK Biobank, monitoring them for dementia over periods of 4 years in the US group and 12 years in the UK group. Over 90% of participants reported drinking alcohol, assessed via questionnaires and the AUDIT-C tool for hazardous patterns like binge drinking.
In observational analyses, a U-shaped association appeared, with non-drinkers and heavy drinkers (40 or more drinks weekly) facing 41-51% higher dementia risk compared to light drinkers (fewer than 7 drinks weekly). However, Mendelian randomization using genetic data from 2.4 million people revealed no such curve. Instead, dementia risk increased steadily with genetically predicted alcohol use.
For instance, an extra 1-3 drinks per week linked to a 15% higher risk, while a doubling in genetic risk for alcohol dependency associated with a 16% increase. The study found no protective effects from low-level drinking, attributing past observational findings to reverse causation: people with early cognitive decline often reduce alcohol intake before diagnosis.
"Our study findings support a detrimental effect of all types of alcohol consumption on dementia risk, with no evidence supporting the previously suggested protective effect of moderate drinking," the researchers concluded. They noted limitations, including stronger associations among people of European ancestry due to sample sizes, and unverified assumptions in genetic methods.
The findings, involving diverse ancestries, urge reducing alcohol as a dementia prevention strategy, emphasizing the need to address reverse causation in prior studies.