Relative and biomarker-based validity of a food-frequency questionnaire estimating intake of fats and cholesterol.

The relative validity of a 104-item food-frequency method to assess intakes of fats and cholesterol was tested against the dietary history of 191 men and women. Pearson correlation coefficients ranged from 0.38 for linoleic acid (% of energy) to 0.83 for energy intake, with 0.78 and 0.75 for the intakes of total fat and saturated fatty acids, respectively. Mean intakes were overestimated by the food-frequency questionnaire relative to the dietary history by 5% for energy and monounsaturated fatty acids (% of energy) and up to 30% for linoleic acid intake. Linoleic acid concentrations in erythrocytes and adipose tissue were used as biomarkers of intake. The correlation of the linoleic acid intake according to the food-frequency questionnaire with linoleic acid in erythrocytes and adipose tissue was 0.44 and 0.28, respectively, and the dietary history gave similar values. The food-frequency questionnaire gives results similar to those from the dietary history and is thus considered appropriate for classifying subjects according to their fat intake.