These notes accompany FoodEdge's October 2026 reporting and practical resources on salt substitutes, AI in food development, continuous fermentation and GLP-1 food-purchase research. They identify the original sources, selected disclosed funding and interests, and the limits of our source checks.
Source-check date: 17 October 2026. A later correction, paper version or regulatory notice can change the account. Source publication dates and study periods are stated separately from the date of FoodEdge's coverage.
US salt-substitute rule
The FDA announcement is dated 9 October 2026. The Federal Register final rule was published on 13 October, with a general effective date of 14 December 2026. The rule provides for a stay of the dairy standards-of-identity provisions to which proper objections are filed. The applicable provision and its current status therefore matter for a particular product.
Source relationship: these are two presentations of the same regulatory action. They are not independent empirical validations of a reformulated food.
Access and limit: the public announcement and relevant Federal Register HTML sections, including Dates, section III.A and section VI, were inspected. The official PDF and every amended standard were not audited. The summary is not a complete product-specific compliance or labelling opinion; ingredient requirements and formulation performance remain separate questions.
Human-AI food-design preprint
Vahidullah Tac and Ellen Kuhl, Generative AI enables human-centered design of sustainable plant-based foods, arXiv:2610.15616. The checked version is v1, submitted 11 October 2026. Read the record and version history and the full v1 paper.
The described workflow combines AI, chef refinement and consumer assessment. The six formulations and 101 participants belong to that combined workflow. The mushroom formulation's reported overall-liking mean of 5.3 on a seven-point scale is compared descriptively with a historical Big Mac reference from an earlier survey. It is not a concurrent head-to-head test establishing sensory equivalence.
Disclosed support: the acknowledgements list a Schmidt Science Fellowship in partnership with the Rhodes Trust for Tac. Support listed for Kuhl includes the Stanford Doerr School of Sustainability Accelerator, Stanford Bio-X Snack Grant, Bezos Earth Fund, NSF CMMI 2320933 and ERC Advanced Grant 101141626. Culinary and sensory collaborators are also acknowledged in the paper.
Interests and access: an explicit competing-interest statement was not established by the targeted check. That is not a conclusion that no interests exist. The version record, methods, results, discussion, limitations and acknowledgements were inspected. Raw data, code and supplementary material were not audited; independent replication and peer-reviewed publication were not established in this check. The HTML and metadata belong to one study, not two independent replications.
ECOFERM continuous-fermentation announcement
The NIZO announcement, ECOFERM Consortium Launches to Advance Sustainable Continuous Fermentation for Alternative Proteins, has a 1 October 2026 date in its body and a 30 September 2026 listing date. Both are retained rather than treated as an industrial-results date.
The announcement describes investigation of continuous fermentation with biomass retention through controlled cell agglomeration. It names challenges for longer campaigns and beta-lactoglobulin as one planned demonstration case. These are a research direction and stated challenges, not independently established industrial performance or cost results.
Disclosed interest and support: NIZO is a participant in the project it describes. The announcement identifies TKI Agri & Food support, project LWV26909. Its account is therefore attributed to a participant, rather than treated as independent industrial validation.
Access and limit: the public announcement and listing were read. No private data room, management interview, campaign dataset, financial model or customer contract was examined. FoodEdge's proposed accepted-output cost boundary is an editorial analytical framework, not a measured ECOFERM result.
Danish GLP-1 food-purchase study
Sørensen and colleagues, Consumer Food Purchases After Glucagon-Like Peptide-1 Receptor Agonist Initiation, JAMA Network Open, 26 January 2026, DOI 10.1001/jamanetworkopen.2025.55449. Read the original article and its disclosures. This is January background evidence used for an October product-development question.
The authors describe a cross-sectional design linking Danish grocery receipts and prescription records, with before-and-after purchase periods and matched comparisons. Nutrient composition is measured per 100 grams of purchased food. This does not measure an individual's total dietary intake or establish the effect of a proposed food product. Limitations include self-selection into the receipt cohort, unavailable BMI data and the wider weight-loss journey.
Selected disclosed funding and interests: Hasselbalch reported a Novo Nordisk Foundation BRIDGE grant during the study. Torp-Pedersen reported grants from Bayer and Novo Nordisk outside the submitted work. These are source disclosures, not a finding that the research is unreliable. Consult the original article for the full funding and interest statements.
Access and limit: the public article's methods, results, discussion and disclosures were inspected during the editorial checks. Supplementary material and raw data were not audited. No registry-wide correction or retraction search was performed. The original journal page should be checked again before relying on the research in a later decision.
How to use these notes
Distinguish a source's findings, a participant's intentions, FoodEdge's interpretation and a proposed next test. Repeated coverage of one source does not add independent confirmation. The worksheets and frameworks in the related FoodEdge materials help structure a decision; they are not presented as field-tested protocols or evidence of a product outcome.
These notes reproduce no third-party figures and grant no new republication licence for source material. Follow the original links for current versions and applicable terms.
Editorial process: prepared with AI assistance from public sources. No external expert review is claimed.