In a 23-year study, long-term moderate alcohol consumption was associated with lower dementia risk.

Prior studies suggest a J- or U-shaped relation between alcohol consumption and incident dementia, with moderate alcohol consumption associated with lowest dementia risk. However, these studies were methodologically flawed (e.g., assessed late-life, rather than lifetime, alcohol consumption).

In this long-term, prospective, U.K. cohort study, researchers assessed the association between alcohol consumption and risk for dementia among 9,100 middle-aged people (age range, 35–55 at study inception) without dementia. Midlife alcohol consumption was based on the mean of three assessments.

After mean follow-up of 23 years, 397 participants had developed dementia. After adjustment for multiple variables, abstinence at midlife was associated with significantly higher risk for dementia than was alcohol consumption of 1 to 14 units weekly.

However, among participants who consumed >14 units weekly at midlife, each 7-unit increase in consumption was associated with a significant 18% increase in relative risk for dementia.