A comparison of frequency and quantitative dietary methods for epidemiologic studies of diet and disease.

Agreement between frequency and quantitative dietary methods was assessed for the ability of frequency intake data to substitute for quantitative intake data in diet-disease investigations. Frequency and quantitative intakes of 342 male subjects participating in an ongoing case-control study in Hawaii during 1981-1982 were obtained using a recall interview method designed to assess usual dietary habits. The extent of agreement between frequency and quantitative intakes of various dietary components (44 food items, 20 food groups, 8 nutrients) was determined at the group and individual subject levels, and with regard to specific study objectives. The results showed that for studies based on aggregate data, frequency and quantitative dietary methods will give reasonably comparable results in analyses involving food items. However, for similar studies involving food groups and nutrients, and for all studies based on individual data, frequency dietary methods cannot be relied on to yield the same diet-disease associations as would corresponding quantitative methods.