A proposed nine-domain nutrient-function architecture for modelling adult nutrient adequacy across dietary preference variants and calorie tiers
Dietary labels, calorie targets and broad food-group advice do not by themselves show whether nutrient functions are preserved when foods are excluded or substituted. This study developed a proposed nine-domain nutrient-function architecture to translate adult nutrient requirements into auditable food-source decisions across contrasting dietary preferences and calorie tiers. The architecture was developed through a design-based, criterion-referenced process beginning with the declared nutrient, dependency, safety and implementation requirement universe; mapping recurrent substitution failures; clustering failures by shared planning action; and applying eight domain-retention criteria. Eleven source-availability archetypes were crossed with four maintenance calorie-tier anchors to generate 44 internal model-evaluation scenarios. Foods consumed periodically were converted to average daily equivalents. Candidate food quantities were iteratively reconciled against tier energy bands, benchmark adequacy, domain coverage, source dependencies, upper-limit screens and practical portions. Compatible ordinary foods were considered first; unresolved gaps or narrow-source vulnerabilities then triggered evaluation of alternative foods, verified fortification, targeted sources or conditional interpretation. The model generated 1,012 non-energy scenario-nutrient rows. Sodium was retained as a safety/contextual output, leaving 22 ordinary adequacy nutrients and 968 ordinary adequacy rows. Final evaluated scenarios met a mean of 20.64 of 22 ordinary nutrients, with a scenario range of 19–21 and variant means of 20.25–21.00. Tier means increased from 20.00 in Lite to 21.00 in High-Output Maintenance. Iron and zinc each produced eight near-adequacy findings; no ordinary nutrient produced a moderate or major gap. Vitamin D remained conditional in all scenarios. Fortified plant milk, fortified nutritional yeast, algal EPA + DHA, iodized salt and a capped selenium source materially affected selected outputs. The safety screen produced 484 no-UL, 374 within-UL, 47 watch and 151 exceedance flags; most exceedance flags involved form-sensitive vitamin A, niacin and folate comparisons or sodium. Thirty-three scenarios showed moderate and 11 high sensitivity to the tested assumptions. The proposed architecture generated broad modelled benchmark coverage while making compensatory sources, conditional findings, source-form safety issues and broader food-matrix pathways visible. These internally constructed scenarios demonstrate feasibility under declared assumptions, not independent validation, clinical efficacy, real-world adherence or universal individual sufficiency.