Reproducibility and relative validity of a newly developed web-based food-frequency questionnaire for assessment of preconception diet

• Published in: BMC nutrition Vol. 5; no. 1; pp. 47 - 12
• Main Authors: Salvesen, L, Hillesund, E R, Vik, F N, Brantsæter, A L, Øverby, N C
• Format: Journal Article
• Language: English
• Published: England BioMed Central 07.11.2019; BMC
• Subjects: Food-frequency questionnaire; Preconception diet; Preconception phase; Relative validity; Reproducibility; Weighed food record; Relative validity; Preconception diet; Weighed food record; Young adults; Reproducibility; Food-frequency questionnaire; Preconception phase
• ISSN: 2055-0928; 2055-0928
• DOI: 10.1186/s40795-019-0310-y

The importance of diet and nutrition during preconception age is a window of opportunity to promote future parental and transgenerational health. As a sub-study to a large Norwegian study, 'Diet today - health of tomorrow', a food-frequency questionnaire (FFQ) was developed to assess diet during the preconception phase in young adults aged 20 - 30 years and in this paper we report the reproducibility and relative validity of this questionnaire. The FFQ was developed from an existing FFQ validated in adolescents. Participants were recruited on social media and at a university. Reproducibility was assessed by comparing the test and retest of the FFQ. Relative validity was assessed by comparing intake measured by the FFQ with a 7-day weighed food record. Energy, nutrients and food intake were used to assess the reproducibility and relative validity of the FFQ. The study applied the Spearman's rank correlation coefficient, percentage of agreement and Cohen's Kappa to assess reproducibility and validity. There were 32 participants recruited to the study, of which 21 participants completed both the test-retest reproducibility and the relative validation. The test-retest reproducibility had a median correlation coefficient of 0.85 for energy and nutrients, a median Spearman's rank correlation coefficient of 0.75 and a median Cohen's Kappa of 0.51 for food groups. The relative validity of the FFQ had a median correlation coefficient of 0.59 for energy and nutrients, a median Spearman's rank correlation coefficient of 0.54 and a median Cohen's Kappa of 0.28 for food groups. This newly developed FFQ for preconception diet in young adults had a satisfactory test-retest reproducibility and fair relative validity.