Calculation of Compound Intake Levels Using Integrated Food Compound Databases and Food Intake Data.

Background: Despite substantial research, the mechanisms through which food influences disease largely remain unclear. Analyzing cohort data at the level of food compounds (i.e. individual molecules present in foods) may reveal new insights, and as a first step, we previously integrated 3 food compound databases. Objectives: This study aimed to combine the integrated food compound databases with food intake data to estimate compound intake levels and evaluate their plausibility. Methods: We used food frequency questionnaire (FFQ) records from 135 adult patients with inflammatory bowel disease from University Medical Center Utrecht. Generic FFQ-based dietary data were translated into compound intake values. We dealt with differing compound values across the databases by applying a systematic prioritization strategy to obtain a single representative value per compound in each food item. Intake amounts of 770 compounds were calculated for all 135 subjects. For vitamins and minerals, plausibility was assessed by checking whether the p50 of our calculated intake fell within the p5 to p95 range of Dutch population intake data. For fatty acids and polyphenols, plausibility was assessed by evaluating whether percentile ranges overlapped with those reported in literature. Results: Our findings indicate that all p50 values of our calculated intake data fell within the p5 to p90 range of the Dutch population intake data for vitamins and minerals. For fatty acids and polyphenols, intake ranges overlapped with those reported in literature for all compounds, but showed more deviation, likely due to regional dietary differences and the absence of standardized population-based intake benchmarks. Conclusions: Our study provides a foundation for food-health research by offering plausible intake estimates for a wide range of dietary compounds derived from Dutch cohort data.