Statistical Analysis
The study protocol and analysis plan prespecified all choices about exposures, outcomes, effect modifiers, and analysis methods . Analyses treated beaches as a fixed effect to derive summary estimates across the 13 cohorts. We estimated cumulative incidence ratios (CIRs) using log-linear models with robust SEs clustered at the household level. Adjusted models controlled for the following characteristics: age, gender, race, reported chronic gastrointestinal problems at enrollment (irritable bowel syndrome, ulcerative colitis), contact with animals, contact with gastrointestinal illness in the 3 days before enrollment, and consumption of undercooked or raw eggs, meat, or fish in the 3 days before enrollment or at any time during follow-up.