Alcohol use and cognition at mid-life: the importance of adjusting for baseline cognitive ability and educational...

The nature of the relationship between cognition and alcohol consumption remains controversial. Studies have reported negative, positive, and nonsignificant effects of alcohol consumption on cognition. Problematic throughout the literature is that baseline cognitive ability has not been adequately controlled in previous studies, and even educational attainment is only sometimes controlled. Because such variables may be associated with both alcohol intake and later-life cognition, we hypothesize that the observed relationship between alcohol intake and cognition may change when these variables or other conditions in early life have been controlled.

We examined the relationship of alcohol intake and cognition at age 53 using the Wisconsin Longitudinal Study, which has followed Wisconsin high school graduates from 1957 to 1992. Our measures include cognitive ability test scores from the freshman and junior years of high school, educational attainment, an abstract reasoning test score at age 53, alcohol intake at age 53, and other measures.

When no controls were used, both men and women with low levels of alcohol consumption at 53 (i.e., 0-1 drink per day) had better scores on the abstract reasoning subtest of the Wechsler Adult Intelligence Scale (WAIS-R) at age 53 than subjects who never drank or currently did not drink. However, after adjusting for adolescent-measured cognitive ability and educational attainment, men with low levels of consumption no longer had higher abstract reasoning scores than nondrinking men, but they still did have higher abstract reasoning scores than men who drank more than one drink per day. For women, adjusting for cognitive ability and educational attainment eliminated all significant effects of alcohol on cognition, and reversed the nonsignificant result that women with higher consumption had the highest cognition scores. These results demonstrate the importance of adjusting for baseline cognitive ability when attempting to study the effect of long-term alcohol use patterns on cognition, and that educational attainment cannot be considered a valid substitute for baseline cognition scores.

Much of the apparent benefit of moderate alcohol intake on cognition in our society may well be explained by differential rates of alcohol consumption among subjects with differing baseline cognitive ability scores. Neither is there evidence that moderate alcohol intake reduces cognitive functioning.