known_biases: single-site African American cohort (Jackson, MS), population-specific FFQ, modest size (~3.5k for incident analysis) — limited power (wide CIs) plus region/ethnicity-specific confounding move all JHS egg estimates together.
known_biases: single-site African American cohort (Jackson, MS), population-specific FFQ, modest size (~3.5k for incident analysis) — limited power (wide CIs) plus region/ethnicity-specific confounding move all JHS egg estimates together.