Inferred vegans in studies not designed to assess vegan status may be largely artifacts

Recent studies from China and India, none designed to assess vegan status, reported unusually large vegan samples (n = ~1000-8000) and prevalences (~1-9%). These were mostly inferred from short food frequency modules or short-term dietary assessment (e.g., 1 month). Many inferred vegans were socioeconomically marginalized or socially excluded. This perspective argues that the reported prevalences largely result from underassessment of intake and misclassification rather than adherence to vegan diets.