Agreement between two dietary methods in reported intake of beer, wine and liquor

This study compares reported beer, wine and liquor intake from a dietary quantity-frequency questionnaire and a 16-day diet diary kept by the same respondents in the 1984-85 University of Michigan Food Frequency Study. The study subjects were 228 black and white men and women, aged 24-51 years. On the two methods, the reported mean ethanol intake derived from each beverage, mean frequency of intake of each beverage and mean quantity for each beverage were similar. The relative rankings of individuals by the amount of ethanol for each beverage were also similar. The methods agreed less well on whether a particular beverage was ever consumed. The absolute amount of ethanol from each beverage agreed more closely between methods than did the percent of ethanol from each beverage. Results were similar for each race-sex subgroup. These findings suggest that analyses should use the reported amount of ethanol from each beverage, rather than converting to percentages or classifying according to the most used beverage. The good general agreement in the types and amounts of alcoholic beverages reported promotes some confidence in the relative validity of data from these two dietary methods for describing moderate alcohol intake in the general population.