The relative validity of a computerized food frequency questionnaire for estimating intake of dietary iron and its absorption modifiers

Abstract

Objective: To determine the relative validity of an iron food frequency questionnaire (iron FFQ) designed to assess intakes of dietary iron and its absorption modifiers.

Design: A computer-administered food frequency questionnaire was designed to estimate intake of total, non-haem, haem and meat iron as well as dietary components which influence iron absorption (vitamin C, phytate, calcium, grammes of meat/fish/poultry, tea and coffee) in women consuming a Western diet. The relative validity of the iron FFQ was assessed by comparing its results with those from weighed diet records collected over 11 days.

Setting: Dunedin, New Zealand.

Participants: Forty-nine women aged 19-31 y attending the University of Otago.

Results: There was good agreement between the iron FFQ and the weighed diet records for median intakes of total iron, non-haem iron, calcium, tea and coffee. For dietary component intakes, correlations between the two methods ranged from 0.39 (for vitamin C) to 0.87 (for coffee) with 0.52 for total iron, and 0.61 for haem iron. In cross-classification with the weighed diet record, the iron FFQ correctly classified between 22% (for vitamin C) and 51% (for phytate) of participants into the same quartile. Actual values for surrogate categories indicated that the questionnaire can clearly differentiate between low and high intakes of all the dietary components assessed. The questionnaire also showed an acceptable level of agreement between repeat administrations (eg a correlation for total iron of 0.65).

Conclusions: The iron FFQ is appropriate for assessing group intakes of total iron, and iron absorption modifiers, in population studies to assess the aetiology and treatment of iron deficiency states in adult women consuming a Western diet.

Sponsorship: This research was supported (in part) by a grant from the Health Research Council of New Zealand.