2026-10-09T19:05:02Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34038117 2026-10-01T04:25:22Z category_374 portal_316 item_type_3 month_year_10_2026

Table 1_Policy themes and instruments in WHO and UNICEF child food-environment guidance, 2015–2025: a computational content analysis.docx Lu Xing (2551372) Xueqi Wu (18275821) Clinical and Sports Nutrition BERTopic child nutrition computational content analysis food environments food governance nutritional health literacy policy discourse policy implementation <p>In children’s food environments, global guidance can shape policy discourse and instrument choice, but its content does not establish implementation or outcomes. In 2015–2025, the World Health Organization (WHO) and the United Nations Children’s Fund (UNICEF) published English-language guidance at the global level. A documented search, 2-author screening, and author-decision-masked expert review produced a corpus of 22 expert-refrozen documents (1,099 pages; 1,259 primary chunks). The BERTopic with document-capped discovery yielded six provisional clusters: food-marketing restrictions and children’s rights; obesity burden; nutrition-sensitive social protection and service delivery; school, physical activity, and healthy food environments; infant and young child food promotion controls; and business engagement and food-industry governance. Among 978 chunks, marketing restrictions and children’s rights accounted for 332 assigned chunks (33.9%). The expert accepted all descriptors and changed three of seven boundary decisions; applying those decisions and rerunning the full pipeline retained six identity-mapped clusters. Fixed-corpus seed and zoning diagnostics were favorable, but context expansion produced seven clusters and dominant-source removal substantially reorganized two source-concentrated clusters. To assess coding reliability, an 84-chunk exercise was archived under the earlier 19-document freeze; issuer and period profiles were descriptive. The study therefore provides a bound, auditable map rather than a stable taxonomy. Nutritional health literacy is a downstream interpretive application: no actual implementation process or literacy, behavioral, dietary-quality, or health outcome was measured.</p> 2026-10-01T04:25:22Z Dataset Dataset 10.3389/fnut.2026.1935052.s001 https://figshare.com/articles/dataset/Table_1_Policy_themes_and_instruments_in_WHO_and_UNICEF_child_food-environment_guidance_2015_2025_a_computational_content_analysis_docx/34038117 CC BY 4.0