To address limited food frequency questionnaire (FFQ) capacity in public health monitoring in Malaysia, we aimed to develop a semi-quantitative FFQ for an adult multiethnic population for comprehensive fatty acid (FA) profiling inclusive of saturated (SFA), monounsaturated (MUFA), polyunsaturated fatty acids (PUFA), PUFA:SFA ratio, trans fatty acids, omega-3 and omega-6 FAs. A 240-food itemed FFQ used diet records (DR) of Malaysia Lipid Study (MLS) participants and fatty acid composition database from laboratory analyzed foods. The developed MLS-FFQ underwent face and content validation before relative validation in a free-living population (n = 114). Validation was facilitated for macronutrient data comparisons between DR and FFQ via Spearman's correlation coefficient analyses; and for fatty acid composition data by independent pairing of DR, FFQ and plasma triglyceride using the triads method. Moderate correlation between dietary methods was obtained for macronutrients and FAs (r = 0.225-0.457, p < 0.05) except for ω-3 FAs, presenting good agreement with grossly misclassified nutrients <10%. For fatty acid composition data, the magnitude of validity coefficients (z) for SFA, PUFA, PUFA:SFA ratios and ω-6 FAs by all 3 methods were not significantly different (p > 0.05). In conclusion, the MLS-FFQ was shown to be a valid tool to assess population dietary intakes.
Keywords: biomarkers; comparative validity; dietary assessment; dietary records; fatty acid composition; food frequency questionnaire.