Performance of a shortened telephone-administered version of a quantitative food frequency questionnaire

PURPOSE: We evaluated the performance of a telephone-administered food frequency questionnaire in a study of 190 men and women, 30–79 years of age, who participated as controls in a study of colon cancer. METHODS: The telephone version of the questionnaire was modified from a longer food frequency questionnaire originally administered in person to each of the participants. One month later, the telephone questionnaire was administered to a subgroup of 190 participants and readministered to 169 members of the subgroup two weeks later to assess the reproducibility and comparative validity of the instrument. RESULTS: The unadjusted correlation for energy between the original in-person full food frequency questionnaire and the abbreviated telephone version was 0.69. The median energy intake from the telephone version was 17% lower in men and 23% lower in women. The energy and sex-adjusted correlation coefficients for other nutrients ranged from 0.45 for vitamin E to 0.78 for fiber. The intraclass correlation coefficients to measure reproducibility ranged from 0.62 for animal protein to 0.83 for folate. CONCLUSIONS: These data indicate that this brief, telephone-administered questionnaire is reproducible and provides a ranking of nutrient intake comparable to that provided by a full in-person interview.