Everything you just read is built on the same evidence hierarchy R3 applies to every topic we cover. We start with primary sources — peer-reviewed studies, regulatory filings (FDA, EPA, CPSC), and standards bodies (NSF, GREENGUARD, OEKO- TEX) — and only then layer in synthesis from credentialed reviewers. Brand whitepapers and marketing copy are weighted near zero. When a finding rests on a single study, we say so. When a study contradicts the prevailing narrative, we surface both sides and tell you which way the evidence actually leans.
For toxicology: nutrition, we prioritize independent toxicology, exposure-pathway research, and verified certification data over anecdote and testimonial. Every external citation in this piece links to a primary source whenever one exists; aggregator summaries are used only when they consolidate data that isn't openly published elsewhere. The goal isn't to give you a closed verdict — it's to hand you the same evidence trail an evidence-literate parent would assemble themselves if they had a free weekend.
R3 is not a medical, legal, or financial advisor. The research summarized here is general consumer-safety reporting, not personalized health guidance. If a finding on this page intersects with a real decision you're making for a child with a known sensitivity, allergy, or medical condition, talk to your pediatrician or a board-certified specialist — they can weigh the evidence against your family's specific situation in a way no article can. We'll update this piece when new credible evidence changes the picture; the “last reviewed” date in the byline is the source of truth on how current this analysis is.
Two more things worth knowing. First: R3 does not accept sponsored placements, paid product reviews, or affiliate- weighted rankings. Every product mentioned in this piece was scored against a category-specific methodology we publish publicly, with the exact same criteria applied to every product in the category. Second: if you spot a citation that has moved, a study that's been retracted, or a methodology gap, the fastest way to flag it is the feedback link in our footer. We treat correction requests as load-bearing — bad citations get pulled, not patched over.
