Indicator foods for vitamin A assessment.

Epidemiological interview studies examining the association between vitamin A and cancer at various sites have been hampered by restricted time available for interview; consequently, studies have included varying lists of food items or broad food groups, thus making comparability of results difficult. To identify a standardized list of indicator foods that adequately assess total vitamin A intake, we examined the 24-hour dietary recall of 13,201 adults who participated in the First National Health and Nutrition Examination Survey (NHANES I) from 1971 to 1974. Food items reportedly consumed were ranked by contribution to overall (aggregated) intake in various subpopulations by an index of vitamin A contribution, which reflected the frequency of consumption, portion size, and vitamin A concentration (IUs/100 g). A comparison of these ranks identified certain food items that had a relative contribution to vitamin A intake which varied by sex/race group, season of interview, age, or region of the country; income level had little effect on the food rankings. The top-ranking 50 foods were sufficient to correctly classify 80%-90% of the individuals into low-, moderate-, and high-consumer categories. The major contributing foods for any subpopulation examined included both retinol (e.g., dairy products, liver) and carotenoid sources of vitamin A (e.g., certain fruits and vegetables) in addition to items (e.g., mixed tomato and cheese dishes) not included in earlier questionnaire studies. Recommendations are made for future questionnaires designed to assess vitamin A.