Recommend individual-level studies to control confounders (e.g., hygiene, healthcare access), propose a randomized trial or controlled feeding study where feasible to test causality, and include microbial outcome measures and blinded laboratory analyses to reduce bias