Abstract
Public health nutrition lacks scalable, objective tools for real-time dietary surveillance. We developed FoodSeq-FLOW (Food Landscape Observation in Wastewater), a genomic platform that sequences chloroplast trnL and mitochondrial 12SV5 DNA in municipal wastewater. Across 183 samples from 21 North Carolina wastewater treatment plants serving 2.1 million people, we detected 184 plant and 116 animal food taxa at a cost of <US $0.01 per person. Wastewater-derived dietary profiles correlated with paired individual stool dietary data (Spearman ρ = 0.64), and 98% of animal-derived sequences mapped to known food taxa. Temporal sampling revealed seasonal shifts in food taxa consistent with regional food availability patterns. Spatial analysis revealed community-level dietary signatures associated with per capita income and education, beer ingredient abundance with discretionary income, tropical fruits and pulses with foreign-born population size, and local seafood with coastal geography. FoodSeq-FLOW extends wastewater-based epidemiology from pathogens to diet, providing a scalable platform that existing global wastewater surveillance networks can deploy to inform nutrition policy and market analytics.