PubMed: Provider: NLM Database: PubMed (National Library of Medicine), https://pubmed.ncbi.nlm.nih.gov Date(s) Searched: October 28, 2026 to October 29, 2026 Date range searched: 2020-present Language: English Publication types excluded: Comment, Editorial, Letter, Case Reports, Preprint Final count: 467 (line 10) Search Terms: No. | Query | Results 1 | "Chocolate"[Mesh] OR "Cacao"[Mesh] | 4,796 2 | chocolate*[tiab] OR cocoa*[tiab] OR cacao[tiab] | 12,683 3 | 1 OR 2 | 13,066 4 | "Child"[Mesh] OR "Adolescent"[Mesh] OR "Pediatrics"[Mesh] | 3,715,810 5 | child[tiab] OR children[tiab] OR childhood[tiab] OR kids[tiab] OR pediatric[tiab] OR paediatric[tiab] OR schoolchild*[tiab] OR "school children"[tiab] OR preschool*[tiab] OR toddler*[tiab] OR adolescen*[tiab] OR teenager*[tiab] OR youth[tiab] OR boys[tiab] OR girls[tiab] | 2,495,095 6 | 4 OR 5 | 4,506,569 7 | 3 AND 6 | 1,786 8 | "Comment"[Publication Type] OR "Editorial"[Publication Type] OR "Letter"[Publication Type] OR "Case Reports"[Publication Type] OR "Preprint"[Publication Type] | 4,884,955 9 | 7 NOT 8 | 1,695 10 | 9 AND english[la] AND 2020:3000[dp] | 467 Final query as run through NCBI E-utilities (line 10): (((("Chocolate"[Mesh] OR "Cacao"[Mesh]) OR (chocolate*[tiab] OR cocoa*[tiab] OR cacao[tiab])) AND (("Child"[Mesh] OR "Adolescent"[Mesh] OR "Pediatrics"[Mesh]) OR (child[tiab] OR children[tiab] OR childhood[tiab] OR kids[tiab] OR pediatric[tiab] OR paediatric[tiab] OR schoolchild*[tiab] OR "school children"[tiab] OR preschool*[tiab] OR toddler*[tiab] OR adolescen*[tiab] OR teenager*[tiab] OR youth[tiab] OR boys[tiab] OR girls[tiab]))) NOT ("Comment"[Publication Type] OR "Editorial"[Publication Type] OR "Letter"[Publication Type] OR "Case Reports"[Publication Type] OR "Preprint"[Publication Type])) AND english[la] AND 2020:3000[dp] AI-use statement: The PubMed search strategy was drafted with the PICO search builder. It used the large language model claude-opus-5-5 (Anthropic; accessed October 29, 2026) to propose search concepts and terms from the review question. The tool made 1 call to the model at high effort (the draft); claude-opus-5-5 was requested and answered it. A refusal fallback was enabled (had the model declined a request, the API could have run it again on another model Anthropic recommends); it was not used. Every MeSH heading, Supplementary Concept and publication type the strategy searches was checked against NLM MeSH (id.nlm.nih.gov) before use: 10 valid. Each numbered line was run in PubMed through NCBI E-utilities between October 28, 2026 and October 29, 2026, and PubMed's own diagnostics (phrases not found, ignored words, lost truncation) were checked. A searcher must review the strategy before it is used. The prompts, the model output and every version are in the provenance file (JSON export). Review question: cocolate impact on kids Run 6ed74a96-cc66-4a1a-8026-ccb04244d1fc, exported October 29, 2026 from the PICO search builder.