Relative Validity and Reproducibility of a Food Frequency Questionnaire to Assess Food Group Intake in Pregnant Jordanian Women

Background: Research on the relationship between maternal diet and pregnancy outcomes requires valid dietary assessment tools in pregnancy.

Objective: This study aimed to develop and examine the relative validity and reproducibility of a quantitative food frequency questionnaire (FFQ) to estimate food group intake in a sample of pregnant Jordanian women.

Design: In this validation study, food group intake from a culturally sensitive quantitative FFQ was compared with food group intake from three 24-hour dietary recalls.

Participants and setting: The validation study was conducted from 2015 to 2017 at the maternity clinics in Jordan University Hospital among 131 healthy singleton pregnant Jordanian women. Of these women, 30 also took part in the reproducibility phase, which involved repeated completion of the FFQ in a time frame of 1 month. Pregnant women who had gestational diabetes, preeclampsia, and chronic diseases were excluded.

Main outcome measures: Relative validity and reproducibility of a 117-item quantitative FFQ used to estimate usual food intake over a period of 1 month.

Statistical analyses performed: Intraclass correlation coefficients and weighted κ statistics were calculated to test the reproducibility between the two administrations of the FFQ. Pearson correlations were estimated to validate the FFQ against 24-hour dietary recalls. Cross-classification and Bland-Altman plots were used to examine the agreement between the two dietary assessment methods.

Results: The intraclass correlation coefficients between the two FFQs ranged from 0.24 for legumes to 0.93 for processed meats. A moderate level of agreement was observed between two FFQs. De-attenuated and energy-adjusted correlations ranged from 0.08 for sweets and sugar to 0.93 for sugary drinks. On average, 50.9% and 45.2% of participants were classified by the FFQ and the 24-hour dietary recalls into the same quartile based on their crude and energy-adjusted food group intake, respectively. Bland-Altman plots showed satisfactory agreement between two methods for most food groups.

Conclusions: The FFQ showed moderate reproducibility and good relative validity for most food groups.