Protocol for the development and validation of an online Fermented Foods Frequency Questionnaire (3FQ) for the assessment of fermented foods consumption patterns across European regions

Background: Fermented foods vary significantly by food substrate and regional consumption patterns. Although they are consumed worldwide, their intake and potential health benefits remain understudied. Europe, in particular, lacks specific consumption recommendations for most fermented foods. Objective: This project, which is under the framework of the Promoting Innovation Of Fermented Foods (PIMENTO) Cooperation in Science and Technology (COST) Action (CA20128), aims to develop a validated tool to quantitatively estimate fermented food intake across 4 European regions. Methods: The Fermented Food Frequency Questionnaire (3FQ) was designed to quantify fermented food intake in terms of frequency and quantity. Fermented foods were categorized into broad groups (eg, dairy, plant-based, meat, beverages) based on product classifications, ensuring that the foods included were genuinely fermented through ingredient analysis according to the International Scientific Association for Probiotics and Prebiotics consensus for fermented foods as a guide. For each main fermented food group, subcategories were determined after detailed discussions by a scientific expert panel that provided country-specific examples. For example, for hard cheeses, Parmigiano was chosen in the Italian version, and Graviera in the Greek version. The questionnaire was developed in English (universal version) and then translated into multiple languages using the back-translation method. Each version was pilot-tested for clarity, and data for the prospective validation were gathered. This included two key steps: (1) assessing repeatability by having participants retake the questionnaire after 6 weeks and (2) confirming accuracy by comparing 3FQ results against 24-hour dietary recalls from a subsample of participants. Statistical analyses will be used to confirm agreement between the methods. Representative sample calculations were performed for 4 groups by biological sex and age group (between 18 and 49.9 years and 50+ years). To ensure ...