Dashboard
Conquer diet-related disease by 2030
A live view of this evidence corpus. It shows what this workspace covers — not disease prevalence, and not an official statistic.
What does conquer mean? — the five-stage definition
- 1 · MAPPED — every major diet→disease edge extracted, pooled, and graded. The corpus sees it.
- 2 · KNOWLEDGE CLOSED — no contested edges; every verdict settled as strong consensus or consensus-on-no-effect; the gap engine prescribes nothing. We know what works.
- 3 · DELIVERING — population intake moving toward the minimum-risk level; the binding constraint is no longer knowledge. People can and do act on it.
- 4 · BENDING — observed incidence tracks the 2030 target pace. The curve responds.
- 5 · CONQUERED — diet-attributable burden held at its floor for 5+ years; any regression revokes the status. It stays down.
Conquer ≠ cure: the claim is bounded by each disease's diet-attributable ceiling — the preventable share prevented, not zero disease. Stages 1–2 are computed live from this corpus (chips on the bench below); stages 3–5 need the tracker's intake and trajectory wiring and stay unclaimed until then. Fail-closed: a stage no data can prove is a stage not shown.
demo rate $1.1T/yr — verify before public usewhere does it go? ▾
Where the $1.1T goes — the ticker, decomposed ▾ hover a slice for its per-day burn and year-to-date · tap → disease rollup
Milken-shaped demo shares of the $1.1T/yr demo rate; the treatment-vs-prevention split is a demo estimate — verify every figure before public use.
Progress toward 2030
Intended impact — which initiatives aim at which diseases, ordered by money spent ▾
Intent mapping is editorial demo — an initiative "aims at" a disease per its stated scope, not proven effect. Money order from the cost strip above. A row with nothing aimed at it is the finding. Verify before public use.
MAHA action tracker — 128 actions, 0 with a published status ▾ September 2025 strategy · each square is one action — a fill requires a published status
Categorization of the 30 food and nutrition actions follows a published legal analysis of the strategy document, not an official breakdown. Square positions carry no meaning — counts only. The empty grid is the finding; fills wire to the accountability tracker when statuses publish.
Coverage by body system
Every association row in the ledger, grouped by the body system its outcome belongs to. Bars share one linear scale and count rows; the number on the right is the studies behind them. This is corpus coverage, not disease burden — a short bar means this workspace has looked less, not that the system matters less.Corpus history
demo trajectory — monthly points shaped to end at today's live totals; real snapshots wire in via the payload manifest (corpus_version · date)
Leading causes of death
age-adjusted deaths per 100,000 · fixed 0–400 scale across years — switch years to watch infectious causes give way to diet-linked chronic disease · each bar splits into ■ diet-attributable and ■ non-diet remainder · totals are CDC/NCHS-shaped, diet shares are GBD-shaped demo PAFs — verify before public use
Disease bench
barometer score per disease — tap to open the rollupStrongest associations
by verdict, then effect — tap to open detailWorkbench — the ingestion plan
general source groups, every one through the same funnel — tap a phase or a kind to filterA public plan, intended for open sharing: every source — a 1942 journal or a 2026 preprint feed — passes the same four-phase funnel, so anyone's corpus can combine and unify here. Volumes are demo estimates; wire to the live extraction queue. The shared payload spec for partner corpora is being drafted.
Latest ingested
auto-extraction queue · scroll for olderAssociation ledger
Select an association
Success Prediction Barometer
composite = 0.40·evidence + 0.30·reduction + 0.30·impact · rolls up live from the ledgerOne level above the association ledger. Evidence = GRADE-weighted by study count (contested ×0.5, sparse ×0.8). Reduction = combined achievable relative reduction over uncontested edges, mediation-discounted so overlapping pathways are not double-counted. Impact figures are placeholder US-prevalence values (millions of adults) — route through the verification queue before public use.
Health outcomes
every outcome in the corpus, grouped by system — groups grow automatically as real data landsClaim language
Structure–function marketing vocabulary — the words on the front of the package, not evidence. Demo set of 120; the full claim database wires in here later. Tap a phrase to filter the ledger by its outcome group.
Outcomes classify into groups by name at render time, so a corpus with dozens of diseases sorts itself; anything unrecognized lands in "Other" instead of disappearing. Tap an outcome to open it in the ledger; "filter ledger" applies the whole group as an advanced filter.
Dietary interventions
evidence-based intervention strategies and their effectiveness · demo records — route to verification queueMediterranean diet
pattern interventionProven cardiovascular benefits and implementation strategies.
Key studies
- PREDIMED · Estruch 2018 · RCT · n=7,447CVD HR 0.70 (0.55–0.89)
- Lyon Diet Heart · de Lorgeril 1999 · RCT · n=605recurrent MI RR 0.28 (0.15–0.53)
- CORDIOPREV · Delgado-Lista 2022 · RCT · n=1,002CVD HR 0.72 (0.54–0.96)
- Sofi 2014 · meta-analysis · pooled cohortsRR 0.90 per 2-point adherence
In the association ledger
DASH diet
pattern interventionEffective hypertension management through dietary approaches.
Key studies
- DASH trial · Appel 1997 · feeding RCT · n=459SBP −5.5 (−7.4 to −3.7)
- DASH-Sodium · Sacks 2001 · feeding RCT · n=412SBP −8.9 with low sodium
- Juraschek 2017 · RCT analysis · n=412−20.8 in baseline SBP ≥150
- Filippou 2020 · meta-analysis · k=30 RCTsSBP −3.2 (−4.2 to −2.3)
In the association ledger
Plant-based diets
pattern interventionComprehensive health benefits and adoption strategies.
Key studies
- Satija 2017 · 3 pooled cohorts · n=209,298T2D HR 0.66 (healthful PDI)
- Qian 2019 · meta-analysis · k=9 cohortsT2D RR 0.77 (0.71–0.84)
- Kim 2019 · ARIC · cohort · n=12,168all-cause HR 0.75 (0.59–0.94)
- Orlich 2013 · AHS-2 · cohort · n=73,308all-cause HR 0.88 (0.80–0.97)
In the association ledger
Study effects are demo values shaped like the literature — verify every number through the verification queue before public use. Intervention patterns enter the ledger as association rows once extracted and gated.
Ultra-processed foods — impact card
experimental · society-behavior application card · one-off exampleOne exposure, all the way down: what it is, how much of the food system it occupies, and what the ledger says it does to health. Society stats are literature-shaped demo endpoints — verify before public use.
What the ledger says
UPF associations in the corpus — tap to open detail
The health-effect list is live from the association ledger; score = 40% settled-share + 30% mean GRADE weight + 30% mean effect size (|ln RR| / ln 2, capped). Society-behavior stats route through the verification queue before any public claim.
Claims evaluator
Matches a claim against the association ledger and returns that edge's graded verdict. It reads only this corpus — it does not search the literature, judge the source, or grade a claim the corpus has not extracted.Paste a health or diet claim
What this does
- 1 · MATCH — parses the claim for an exposure and an outcome, then finds the association rows that carry both.
- 2 · DIRECTION — reads whether the claim asserts benefit or harm, and compares it with the pooled direction.
- 3 · VERDICT — returns that edge's agreement verdict, pooled effect, GRADE certainty, and study count.
- 4 · GAP — if the evidence is thin or contested, names the study the gap engine says would settle it.
- 5 · MECHANISM soon — Show the science will walk the edge's mechanism chain from the Biology as Code pathway set (26 principles ride in the payload today; the walk is not built yet).
Fail-closed: a claim with no matching edge returns NOT IN CORPUS, never a guess. Verdicts are the ledger's, so they carry the same demo caveats as every other view.
Standardization by layer in nutrition science
nine layers, 0–8 · share of each layer both standardized and in routine useScores are calibrated judgments, not measurements — each blends the maturity of available standards with their observed adoption in published data and literature. The audit's master finding: maturity and adoption diverge sharply — excellent standards exist and go unused. Hover a row for the one-line why. Sources: research_/standardization-by-layer.pdf · “Nutrition Science Standardization: A Nine-Layer Stack Audit and Recalibration.”
The ○ benchmark rows come from the eleven-field scoreboard in research_/standard_study.md, toggleable above. Metrology proper (BIPM / ISO 17025) is the physics pole — the corpus scores no field named “physics”; metrology is its measurement regime (the book's rule: “the closer a claim sits to chemistry and physics, the more standardized it is”). Clinical trials / regulatory pharma is the cross-field study's recommended primary analogue for nutrition; chemistry and astronomy are explicitly flagged there as flattering nutrition. All benchmark scores are calibrated judgments and defensible relative rankings, not precise measurements — ordinal, not interval. The dashed line sits at the selected benchmark's weakest layer — an editorial threshold; no document in the repo defines a numeric “standardized” cutoff. Layer 9 has no comparator scores.
Baseline note — canonical as of 2026-08-15. Four inconsistent nine-layer baselines circulated in the corpus. Per the Stack Audit's Stage-2 recommendation, the canonical score set is the audit recalibration (55 · 60 · 33 · 30 · 12 · 35 · 60 · 8 · 8 + linkage ~15) — the default view above, and the calibration every cross-field comparison is stated against. The “as charted” set (60 · 65 · 35 · 45 · 15 · 40 · 55 · 5 · 10, research_/standardization-by-layer.pdf) is retained only as the superseded book-figure estimate.
Nutrition science pipeline — careers
who produces the field's people — the adopters every layer above depends onPhase 0 → 6: the U.S. route from first exposure to specialist or scientist. Standards don't adopt themselves — the people this pipeline produces do.
Phase 0
Explore the field
1–6 months
Shadow RDNs; decide clinical practice vs research scientist track early.
Phase 1
Degree + DPD
2–4 years
ACEND-accredited coursework — biochem, physiology, food science, stats.
Phase 2
Graduate degree
1–3 years
Required for the RDN since 2024 — M.S., MPH, or research-heavy program.
Phase 3
Supervised practice
6–12 months
1,000+ DI hours across clinical, community, and food-service rotations.
Phase 4
Exam + licensure
1–3 months
CDR national exam, then state licensure (LD / LDN) where required.
Phase 5
First role
1–3 years
Build depth in one specialty; document outcomes for specialty boards.
Phase 6
Specialize or research
ongoing
CSSD, CDCES, CSO boards — or a Ph.D. into the scientist track.
Annual output — IPEDS 2021–22, first majors
The pipeline narrows ×39 from bachelor's to research doctorate — the people who will build and adopt the field's standards number 222 a year. Bars share one linear scale; the sliver is the finding.
Where the pipeline leads — career roles
Clinical RDN
bedside & clinic care
Nutrition Scientist
lab, cohort, or industry R&D
Food Industry Nutritionist
product & claims
Public Health Nutritionist
programs & populations
Private Practice / Consulting
independent client work
Faculty / Educator
teach the next generation
Science Writer / Media
translate evidence publicly
Policy & Advocacy
laws, labels, systems
Card data from the book's nutrition-ecosystem hub (book/tools/nutritionecosystem.ts) and appendix A.18 “Where the Degrees Come From” (IPEDS C2022_A_RV). Census note: the largest single bachelor's producer is an online unit, and the top two master's producers are chiropractic colleges — the people pipeline is no more standardized than the data. Verify before public use.
The definition register — terms the system runs on
16 terms · four states · chips open the federal documentsUltra-processed food
Reduction directed across federal child nutrition programs. Joint FDA/USDA request for information, docket FDA-2025-N-1793, 25 July 2025 (90 FR 35305); comment period extended to 23 October; 5,136 comments received. The notice itself states that no single universally accepted definition exists. A document sits at OMB under RIN 0910-ZD60 as a white paper, not a rule — no legal deadline.
Highly processed food
Used in place of “ultra-processed” in the 2025–2030 Dietary Guidelines, released 7 January 2026, which direct Americans to avoid highly processed packaged, prepared, and ready-to-eat foods. Also not federally defined — one undefined term substituted for another.
Natural
FDA requested comments 28 December 2015; the period closed 10 May 2016 with 4,148 comments; no proposed rule followed. FDA operates a policy statement rather than a regulation. Among the most common claims on new U.S. food products — and among the most litigated.
Real food
The tagline of the 2025–2030 Dietary Guidelines and the name of the domain hosting them. No definition anywhere in federal regulation.
Nutrient-dense
Used throughout the Guidelines, including in the protein recommendation. Described but not defined quantitatively. Competing profiling systems — NRF, Nutri-Score, Health Star Rating — rank the same foods differently.
High-quality protein
Appears in the Guidelines' central recommendation. DIAAS and PDCAAS exist as scoring methods; neither is the basis of a federal definition of the term as used.
Refined carbohydrate
Appears in the same recommendation. No federal definition.
Whole grain
FDA draft guidance issued 2006, never finalized. USDA operates a separate whole-grain-rich standard for school meals. A third definition is maintained privately by a trade council.
Clean label
Commercial only — no definition in any jurisdiction. Listed because it drives reformulation decisions at scale.
Dietary fiber
FDA established a labeling definition in the 2016 Nutrition Facts rule; Codex modified its definition in 2009. The definition exists — yet multiple validated analytical methods still return systematically different values for the same food, because they count different components. See Layer 1 — Method, above.
Protein
Defined operationally as total nitrogen multiplied by a conversion factor. The factor is food-specific, several remain under revision, and FAO/WHO revisited them for soy and milk ingredients as recently as 2020. A definition that specifies a proxy.
Unassignable
A formal evidence grade, distinct from limited — the evidence could not be graded at all. The 2025 Dietary Guidelines Advisory Committee applied it to the ultra-processed-food evidence for various populations. Thirteen months later the Guidelines issued a population-wide directive on that category. The grade lives in an advisory report; the directive lives in policy; no mechanism connects them.
Healthy
Regulated since 1994. The original criteria excluded nuts, salmon, avocados, olive oil, and eggs while permitting sweetened cereals. Proposed rule 29 September 2022; final rule 19 December 2024; effective 25 February 2025; compliance 25 February 2028. Roughly nine years from trigger to final rule — thirty-four from the original definition.
Serving size
Defined by FDA as Reference Amounts Customarily Consumed. Defined, enforced, and not in dispute.
Food insecurity
USDA maintains an 18-item measurement module with defined thresholds — the clearest case in nutrition of a construct successfully operationalized. Collection was terminated in September 2025.
Register from the book's data audit — dockets, dates, and comment counts as recorded there; route through the verification queue before public use. External chips open regulations.gov, the Federal Register, eCFR, FDA, USDA, and ODPHP.