The role of portion information in the agreement between food frequency questionnaires and food recalls in older women

Food frequency questionnaires (FFQs) are widely used to estimate usual intake. A questionnaire without portion sizes may facilitate data collection. We examined the nutrient agreement between three days of food recall (FR) and FFQ in older women when portion information is used or ignored. When mean values were compared to FR, there was no consistent bias when using or ignoring FFQ portions. The correlation between either FFQ calculation and FR was significant for most nutrients, and 53% of the subjects were perfectly classified into categories of dietary intake. Thus, based on the data omitting FFQ portions results in overall small differences in the data obtained when portions are used.