Gaps in epidemiologic research methods: design considerations for studies that use food-frequency questionnaires.

It is increasingly common for analytic epidemiology studies of diet and disease to select a semi-quantitative food-frequency questionnaire for dietary assessment. Reasons include its low cost and focus on usual intake. However, the components of variation in nutrient intake based on such methods are poorly understood, and it does not provide a quantitative estimate of nutrient intakes. Because it is unlikely that such methods will be improved substantially, studies using a food-frequency questionnaire should include a dietary standardization substudy. A substudy, conducted in a sample of study participants and aimed at collecting multiple days of diet recall or record, provides data for quantitative calibration of frequency-derived nutrient intake estimates and adjustment of risk estimates for measurement error. Attention should also be directed at other aspects of study design, such as study population selection to maximize dietary exposure contrasts, that can increase the informativeness of epidemiologic studies.