A valid semiquantitative food frequency questionnaire to measure fish consumption.

OBJECTIVE: To validate a detailed semiquantitative food frequency questionnaire designed to measure habitual fish and seafood consumption.
DESIGN: Cross-sectional validation study using an independent biomarker of fish consumption.
SETTING: Perth metropolitan area, Western Australia.
SUBJECTS: Ninety-one healthy volunteers of both sexes aged 21-75 years.
METHODS: Participants completed the questionnaire and provided a fasting blood sample for erythrocyte membrane omega-3 fatty acid (eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)) analysis. The questionnaire was then validated by linear regression analysis of EPA and DHA levels on categories of fish and seafood and overall consumption, adjusted for age, sex, smoking status, body mass index (BMI) and alcohol intake.
RESULTS: Regression coefficients were statistically significant for most fish and seafood items with both EPA and DHA. The strongest association was observed between oily fish and EPA, whereas no significant association was observed between lean fish and omega-3 fatty acids. Variation in omega-3 fatty acids was best accounted for by a model containing variables representing different categories of fish and seafood consumption (R (2) 0.484), rather than a single variable representing overall fish and seafood consumption (R (2) 0.313).
CONCLUSIONS: This study confirms that the varying content of omega-3 fatty acids in foods are reflected in omega-3 biomarkers, and that the questionnaire is a valid measure of fish consumption that enables differentiation between cooking and processing methods and oily versus lean fish intake.