Validation of a food frequency questionnaire for pregnant Swedish women in the NorthPop Birth Cohort Study against repeated 24-hour recalls.

BACKGROUND

Diet during pregnancy significantly influences maternal and offspring health outcomes, including long-term risk of obesity, cardiovascular disease, asthma, allergic disease and diabetes. Food frequency questionnaires (FFQs) are cost-effective tools for dietary assessment but require validation to ensure accuracy. The objective of this study was to assess the validity of an FFQ developed for pregnant women in the NorthPop Birth Cohort Study (NorthPop) in Sweden.

METHODS

Pregnant women (n = 96) were recruited from NorthPop. Nutrient and food group intakes assessed by the FFQ were compared with intakes assessed by three 24-hour dietary recalls (24HR) using group mean comparisons, Spearman correlation coefficients, cross-classification, and Bland-Altman analysis.

RESULTS

The FFQ showed acceptable to very good correlations for most nutrients and food groups. Mean deattenuated correlation coefficients were 0.45 (0.15 to 0.73) for nutrients and 0.55 (0.32 to 0.80) for food groups. Energy adjustment altered individual nutrient correlations, but the overall mean remained unchanged at 0.45 (0.11 to 0.84); however, it reduced the mean correlation for food groups to 0.49 (0.17 to 0.70). Analyses of mean intake, cross-classification, and Bland-Altman plots showed acceptable agreement between the FFQ and the 24HR reference method.

CONCLUSIONS

The FFQ evaluated in this study is a valid tool for assessing dietary intake at the group level among pregnant women in the NorthPop cohort. The results support the use of this FFQ to study relationships between diet and disease and to identify dietary factors during pregnancy.