Validity and Reproducibility of a Semiquantitative Food Frequency Questionnaire and Food Picture Book in Nigeria

Abstract

BackgroundIncreasing noncommunicable diseases in Nigeria are partly related to dietary factors. However, the lack of validated nutrition assessment tools hinders the conduct of nutritional epidemiology research in this population. ObjectivesTo develop a Food Frequency Questionnaire (FFQ) and Food Picture Book (FPB) for Nigerian adults, and to assess its reproducibility and validity compared with 24-h dietary recalls (24DRs) during different seasons in the year. MethodsWe compiled 202 foods for the FFQ through focus groups and consultations with local dietitians. We created an FPB with standardized food portion images to enhance the accuracy of reports of dietary intakes. We administered the FFQs to 205 randomly selected adults in Ibadan, Nigeria at ∼6 monthly intervals between November 2018 and October 2020. We evaluated the FFQ’s reproducibility and validity compared with 24DR across the dry and rainy seasons by examining the consumption of common food and mixed dishes. We computed the Spearman’s correlation coefficients (SCC), intraclass correlation coefficients (ICC), and Wilcoxon signed-rank tests, and generated Bland and Altman plots. ResultsOverall, we studied 110 women (53.7%) and 95 men (46.3%) with a mean age of 45.0 ± 13.4 y (mean ± SD). The reproducibility tests showed a mean ± SD SCC of 0.39 ± 0.14 and mean ± SD ICC of 0.32 ± 0.12. Higher mean ± SD SCC values were noted for cereal products (0.43 ± 0.09), starchy roots and tubers (0.45 ± 0.17), and soups (0.44 ± 0.20). Conversely, lower mean ± SD SCC values were observed for milk products (0.29 ± 0.02), solid fats (0.29 ± 0.26), and fish (0.22 ± 0.19). Regarding validity tests, the overall mean ± SD SCC was 0.27 ± 0.16 and mean ± SD ICC was 0.26 ± 0.16. We observed seasonal variations in intakes of fruits, cassava flour-based products, and nuts, although most foods did not show significant differences in intakes between seasons. ConclusionsOur FFQ and FPB demonstrated moderate correlations and seasonal variations in intakes of certain foods, emphasizing the need to account for seasonality in dietary intakes in nutritional studies in Nigeria and similar countries.