Assessing Alcohol Consumption in Epidemiological Studies

Alcoholic beverage consumption was assessed at each exam using the CARDIA Alcohol Use Questionnaire (AUQ) that queried participants on annual, monthly, weekly and daily alcoholic beverage intake. Alcoholic beverage consumption was defined based on the following questions: “Did you drink any alcoholic beverages in the past year?”; “How many drinks of wine (5 oz glass) do you usually have per week?”; “How many drinks of beer (12 oz glass) do you usually have per week?”; “How many drinks per week do you usually have of hard liquor (1 1/2 oz)?” To our knowledge there is no validated questionnaire available to measure usual alcoholic beverage intake in epidemiologic studies. However, in exploratory analyses, we found that alcoholic beverage intake (milliliters of alcohol per day) estimated from the AUQ was strongly correlated with mean alcohol intake (grams per day) estimated from the validated CARDIA diet history at exam years 0 (r = 0.77; p < .001); year 7 (r = 0.71; p < .001); and year 20 (r = 0.79; p < .001).
To describe the distribution of socio-demographic and lifestyle characteristics of people who drank compared to those who didn’t drink at baseline, participants were categorized into sex-specific drinking categories using alcoholic beverage intake at exam year 0. Category definitions were based on the National Institute on Alcohol Abuse and Alcoholism (NIAAA) guidance on drinking levels [34 (link),53 (link)–56 (link)]. Based on the sum of the usual intake of beer, wine, and liquor/mixed drinks per week (drinks/wk) as reported on the AUQ at exam year 0, men were classified as “non-drinker”, “light drinker” (<7 drinks/wk), “moderate drinker” (7 to 14 drinks/wk), or “excessive drinker” (>14 drinks/wk), and women were classified as “non-drinker”, “light drinker” (<4 drinks/wk), “moderate drinker” (4 to 7 drinks/wk), or “excessive drinker (> 7 drinks/wk) [8 ,53 (link)].