Secondary analysis of a marketing research database reveals patterns in dairy product purchases over time.

OBJECTIVE Development of a method using marketing research data to assess food purchase behavior and consequent nutrient availability for purposes of nutrition surveillance, evaluation of intervention effects, and epidemiologic studies of diet-health relationships. DESIGN Data collected on household food purchases accrued over a 13-week period were selected by using Universal Product Code numbers and household characteristics from a marketing research database. Universal Product Code numbers for 39,408 dairy product purchases were linked to a standard reference for food composition to estimate the nutrient content of foods purchased over time. SUBJECTS/SETTING Two thousand one hundred sixty-one households located in Victoria, Texas, and surrounding communities who were active members of a frequent shopper program. ANALYSES Demographic characteristics of sample households and the nutrient content of their dairy product purchases were analyzed using frequency distribution, cross tabulation, analysis of variance, and t test procedures. RESULTS A method for using marketing research data was successfully used to estimate household purchases of specific foods and their nutrient content from a marketing database containing hundreds of thousands of records. Distribution of dairy product purchases and their concomitant nutrients between Hispanic and non-Hispanic households were significant (P<.01, P<.001, respectively) and sustained over time. APPLICATION/CONCLUSIONS Purchase records from large, nationally representative panels of shoppers, such as those maintained by major market research companies, might be used to accomplish detailed longitudinal epidemiologic studies or surveillance of national food- and nutrient-purchasing patterns within and between countries and segments of their respective populations.