Dietary intake is tightly coupled to gut microbiota composition, human metabolism, and to the incidence of virtually all major chronic diseases. Dietary and nutrient intake are usually quantified using dietary questionnaires, which tend to focus on broad food categories, suffer from self-reporting biases, and require strong compliance from study participants. Here, we present MEDI (Metagenomic Estimation of Dietary Intake): a method for quantifying dietary intake using food-derived DNA in stool metagenomes. We show that food items can be accurately detected in metagenomic shotgun sequencing data, even when present at low abundances (>10 reads). Furthermore, we show how dietary intake, in terms of DNA abundance from specific organisms, can be converted into a detailed metabolic representation of nutrient intake. MEDI could identify the onset of solid food consumption in infants and it accurately predicted food questionnaire responses in an adult population. Additionally, we were able to identify specific dietary features associated with metabolic syndrome in a large clinical cohort, providing a proof-of-concept for detailed quantification of individual-specific dietary patterns without the need for questionnaires.
Our latest preprint is @thaasophobia's magnum opus. Basically, it's a method for inferring dietary and nutritional intake from food-related DNA present in human stool.
— Sean Gibbons 🦠💩 @gibbological.bsky.social (@gibbological) February 6, 2024
Metagenomic Estimation of Dietary Intake (MEDI)
💩🧬➡️🍎🥦🥩🐟https://t.co/8rlfzBU52U@isbsci pic.twitter.com/wOoVHDepK2
Happy to announce that my last preprint from the Gibbons Lab at ISB is now out at https://t.co/9EAybKLLMT . In the end it took us almost 5 years to make this work well.
— Christian Diener (@thaasophobia) February 6, 2024