# Relative Validity and Reproducibility of a Semi-quantitative Food Frequency Questionnaire to Assess Nutrient Intakes of PERSIAN Cohort Participants: Comparisons to 24-Hour Dietary Recalls and Selected Biomarkers

## TL;DR

This study tested a food frequency questionnaire used in the PERSIAN Cohort to see how well it measures nutrient intake compared to 24-hour recalls and biomarkers.

## Contribution

The study provides validation and reproducibility data for a semi-quantitative FFQ in the PERSIAN Cohort using 24-hour recalls and biomarkers.

## Key findings

- The FFQ showed moderate to high correlations with 24-hour dietary recalls for energy and macronutrients.
- Biomarker comparisons showed acceptable validity for urinary protein, sodium, serum folate, and fatty acids.
- Reproducibility was high for most nutrients, with 19 out of 30 showing strong correlations between FFQ1 and FFQ2.

## Abstract

Food frequency questionnaires (FFQs) are widely used in epidemiologic studies to assess the dietary intake of individuals. In this study, we evaluated the validity and reproducibility of the FFQ used in the Prospective Epidemiological Research Studies in IrAN (PERSIAN) for assessing nutrient intakes.

Individuals (n=978) from seven PERSIAN cohort centers participated in this study; an initial FFQ was completed for each person upon enrollment (FFQ1), followed by two 24-hour dietary recalls (24HR) each month, for twelve months, and finally, another FFQ at the end of the study (FFQ2). Serum and 24-hour urine (24H-U) samples were also collected each season, and selected biomarkers were measured. To assess validity, correlation coefficients between the 24HRs and the FFQs were obtained. The triad method was used to compare the biomarkers to the FFQs. Correlations between FFQ1 and FFQ2 were evaluated to assess the reproducibility of the questionnaire.

Correlations obtained for energy and macronutrients, in comparing FFQ1 and FFQ2 to the 24HRs were 0.57,0.63 (energy), 0.56,0.62 (protein), 0.51,0.55 (lipids) and 0.42,0.51 (carbohydrates), respectively. Moderate (0.4‒0.6) and high (>0.6) correlations were seen for micronutrients, with only vitamins B6 and B12 being poorly correlated (<0.4). Validity coefficients obtained for urinary protein and sodium, as well as serum folate and selected fatty acids were acceptably above 0.4. Reproducibility correlations ranged from 0.18 (alpha-linoleic acid) to 0.78 (selenium), with 19 of the 30 evaluated nutrients showing high and 2 showing poor correlations.

Overall, the PERSIAN Cohort FFQ is acceptable to rank individuals based on their nutrient intakes.