Two cliffs, not one: bidirectional nutrient targeting, the supplement blind spot, and the limits of current diet-quality assessment

Diet-quality assessment is examined for whether it represents the two-sided nature of nutrient–health relationships. Risk rises at both ends of the intake range for many micronutrients, yet diet-quality indices score components monotonically and exclude supplements by construction, while population reference values are used as individual targets despite guidance against this. In this critical narrative review, literature was identified through PubMed/MEDLINE, Scopus and the Cochrane Library to September 2026, with direct consultation of the reference-value documents of the National Academies of Sciences, Engineering, and Medicine, the relevant United Kingdom expert committees, and the European Food Safety Authority; selection was purposive rather than systematic. Seven vitamins have no upper limit because adverse-effect data are insufficient, not because safety is shown, yet effects of excess are documented for most micronutrients. Where limits exist, the gap between requirement and limit is three- to seven-fold for vitamin A, zinc, iron and selenium. Supplements supply one to three orders of magnitude above requirement and drive most exceedance in adults; fortification does in children. Of eight instruments in routine use, none applies a two-sided rule at nutrient level or counts supplements. A candidate module is set out covering every micronutrient, anchored on the estimated average requirement and the upper limit, separating food, fortification and supplements, and weighting penalties across four evidence tiers, together with the staged work that would establish whether it is worth adopting.