Relative validation of a food frequency questionnaire to assess dietary fatty acid intake

Abstract

Objectives: This study aimed to develop and validate a Food Frequency Questionnaire (FFQ) for assessing consumption of fatty acids among pregnant women. Methods: Two lists of foods were created according to percent contribution of each nutrient estimated by three 24-hour recalls: a long and short version FFQ to estimate long-chain polyunsaturated fatty acids (LCPUFAs). Student paired t-test and Pearson correlation coefficients were used to verify the differences in mean consumption of nutrients from the FFQ and 24-hour recall. The concordance between the consumption values of the two methods was assessed using the Bland-Altman method and quartiles concordance. Results: For the FFQ-long version, correlation values ranged from 0.33 (p<0.05) to 0.62 (p<0.01) for docosahexaenoic acid (DHA) and linoleic acid (LA), respectively. Eicosapentaenoic acid (EPA) and docosapentaenoic acid (DPA) were not correlated. Exact concordance ranged from 49.0% (energy) to 22.4% (EPA), and discordance ranged from 14.3% (DPA) to 2.0% (Saturated). The FFQ-short version had high correlations for LCPUFAs. Exact concordance ranged from 36.7% (n-3 LCPUFA) to 16.3% (DHA); and discordance from 12.2% (DPA) to 2.0% (arachidonic acid). Bland-Altman analysis showed good concordance for both versions. Conclusion: This nutrient-specific FFQ is a valid instrument to be used to estimate the level of consumption of fatty acids among pregnant women.