Validation of a quantitative FFQ for a study of diet and risk of colorectal adenoma among Japanese Brazilians.

Abstract

To assess the validity of a 161-item quantitative FFQ (QFFQ) that was developed to evaluate dietary risk factors for a colorectal adenoma case–control study. A cross-sectional validation study of the QFFQ against 4 d food diary using Pearson correlation coefficients, cross-classification, weighted k statistics and Bland–Altman plotting. Two hospitals in Sa˜o Paulo, Brazil. Ninety-seven healthy Japanese-Brazilian adults (40–75 years) were recruited. One participant was excluded from the analysis due to unusual energy intake report. Mean daily nutrient intakes from the QFFQ were higher than from the food diary. The mean Pearson correlation coefficient for nutrient intakes between the QFFQ and the average of the 4 d food diary was 0?43, and increased to 0?45 after correcting correlations for attenuation due to residual day-to-day variation in the food diary measurements. Adjustment for total energy and further adjustment for age and gender decreased the correlation; however, 77% of observations remained in the same or adjacent quartiles with a mean weighted k of 0?22. Bland–Altman plots on loge-transformed data showed no linear trend between the differences and means for energy, fat, protein, total folate and vitamin C. Compared with the food diary, the QFFQ showed consistently reasonable performance for dietary fibre, total folate, retinol, riboflavin and vitamin C. This investigation supports the relative validity of the QFFQ as a method for assessing long-term dietary intake. The instrument will be a useful tool in the analysis of diet–adenoma associations in the case–control study.