Methods for analysis of a food frequency questionnaire

Fifty-five students and staff at Robert Gordon's Institute of Technology completed a food frequency questionnaire and then weighed their food for one week. Several methods of calculating nutrient intakes from the questionnaire are compared to results from the weighed inventory. Intakes of energy and several nutrients calculated using estimated typical portion weights were less than those derived from the weighed inventory. Food groups contributing to the apparent under-estimation have been identified. A further method of calculation involved derivation of a factor for each food group such that the frequency estimated from the questionnaire multiplied by the factor equalled the average weight of the foods in that group that had been eaten. Using these factors instead of the typical portion weights resulted in closer agreement to the weighed inventory method, but the difference ranged from positive for some groups to negative for others. Use of separate factors for each group of subjects reduced the range of differences to the weighed inventory method.