Content and face validity of a food frequency questionnaire for the assessment of ultra-processed food consumption in the Italian adult population: a pilot study

Background – The multidimensional complexity of dietary intake, coupled with the current uncertainty around the Ultra-Processed Foods (UPFs) concept, underscores the need for tailored, valid, and reliable dietary assessment tools. This pilot study assessed face and content validity of the novel modified NOVA Food Frequency Questionnaire (mNOVA FFQ), designed to assess UPF consumption among Italian adults, to finalize a version for subsequent testing. Methods – A sample of 52 Italian volunteers from two institutions was recruited. Participants self-completed an initial mNOVA FFQ version, followed by an online/in-person semi-structured interview. A trained interviewer administered 12 questions evaluating understanding/interpretability of the mNOVA FFQ, and a 24 h dietary recall (24HR) to better characterize eating habits in the target population and inform potential refinements to the food list. Both quantitative and qualitative analyses were performed to evaluate the content and face validity of the FFQ. Results – Twenty FFQ items exhibited predominantly negative responses (>50%) to the main questions. All foods reported in the 24HRs were represented in the FFQ, indicating comprehensive coverage of participant diets. Thirteen FFQ items were not consumed in the 24 h prior to the interview, and four of these were consumed by less than 1% of Italian adults, according to national data. Overall, participants perceived the questionnaire as clear and generally easy to understand, although thematic analysis of their feedback highlighted several areas for improvement in the FFQ. Conclusion – This study enabled refinement of the questionnaire closed list of food items and structure, and enhance its clarity and understandability, underscoring the importance of evaluating both content and face validity, and involving the target population in the development of new tools. Further investigation is required to establish the reliability and criterion validity of the mNOVA FFQ for its application within the Italian adult population.