This is your food, drawn honestly — for your body, not an average person.
A blank stays blank — never drawn as zero.
Where science doesn't know a number, the cell says so.
⬡ bFood Hexagon
your nutrients, drawn honestly, never zero
for your body, not an average
40 cells = the NASEM/IOM DRI set (EER via the 2023
equations, protein 0.8 g/kg, nine EAAs in mg/kg — the only body-scaled rows) against USDA
FoodData Central composition, per record ID, refetchable. Three render states, never
collapsed: measured (solid rim) · NotMeasured (dashed, empty, n/m — never zero) · floor
(dashed, filled, ≥ — a sum over foods missing a published row is a lower bound wearing a
total's name). Every requirement cites the authority that set it; recompute the page from
its sources and catch us.
What if the plate told the truth — including the blanks?
A blank that stayed blank is reverence — not a zero.
A blank stays blank — never drawn as zero.
Where science doesn't know a number, the cell says so.
A blank stays blank — never drawn as zero.
Seven cells. Some filled. One blank on purpose.
Not measured
1 · YOUR BODY — the five inputs that genuinely move a requirement
How these values are used
Only these four scale with your body, and this page
will not pretend otherwise. Energy uses the NASEM 2023 EER equation, where your height,
weight, age and activity are the variables. Protein is an RDA in grams per kilogram, so
your mass is the number. Fibre scales through energy at 14 g per 1,000 kcal.
Every vitamin and mineral below is a flat constant for your age and sex group — NASEM
uses reference weights inside the committee's derivation, never as a coefficient in the published
value. Inventing a weight curve for zinc to make this page feel more personal would be the exact
fake precision this project exists to refuse, so the cells say 👥 group
and mean it.
—
2 · THE HEXAGON — 40 cells, one per nutrient, filled by what you actually got
3 · THE BASKET — three honest weights, never one implied
The house serving is 1.25 g of hemp hearts per kg of
body weight — 100 g for an 80 kg adult, founder-set, and labelled a house basis rather
than a published standard, because it is one. It moves when your weight moves. The reference
intakes beside it do not. Two numbers that share a unit but not a pedigree must never be
allowed to touch.
4 · THE LEDGER — every nutrient, the exact amount, and where it came from
Sorted by how well the basket covers it. The tick on each bar is 100% of your
requirement. ≥ means a food in your basket has no published row for that nutrient, so the
total is a floor, not a total — the same law that forbids a lab's four-analyte panel from
calling itself "total cannabinoids".
| nutrient | coverage | % | you need | basket gives | basis | gap |
|---|
5 · PROTEIN QUALITY — why 31.6 grams of protein is not 31.6 grams of protein
This is the fourth time this project has caught the same error class,
and this time it is caught on our own favourite food. A lab's "total cannabinoids" is a sum
over four analytes wearing a total's name. "Saturated fat" is two things a body routes
differently, collapsed into one row. A basket total is a floor when a food has no published row.
And a food's protein grams are an upper bound on usable protein, not a measurement of it.
THE MECHANISM — established biochemistry, not an outcome claim
A ribosome reading an mRNA needs the exact amino acid its codon calls for, at the moment it reads it. There is no substitution and no skipping. So the number of complete proteins a meal can build is capped by whichever indispensable amino acid runs out first relative to need — the limiting amino acid. And humans have no storage depot for amino acids: nothing like glycogen for glucose. The surplus cannot be banked. It is deaminated in the liver, its nitrogen fixed into urea and excreted, and its carbon skeleton burned or stored as fat. The excess still delivers calories. It delivers no protein.
AND THE BRAKE, which matters more than the finding
The limiting amino acid is a property of the whole day's amino acid supply, not of any single food. The Academy of Nutrition and Dietetics, current position: "The terms complete and incomplete are misleading in relation to plant protein. Protein from a variety of plant foods, eaten during the course of a day, supplies enough of all indispensable amino acids when caloric requirements are met." During the course of a day — not at the same meal. That "combine at every meal" rule was retracted decades ago. So this page will never print an "incomplete protein" warning beside a food: it would be mechanically derived and substantively false.
Hemp's first limiting amino acid is lysine — and
it is lysine for every cereal and seed protein, while it is methionine+cysteine for every legume.
That mismatch is exactly why complementation works: two proteins limited by different
things, pooled, are limited by neither. It is not a trick, it is arithmetic on the aggregate.
✓ AND THE HEXAGON REPRODUCES IT INDEPENDENTLY — which is the only reason you should believe either
The claim above comes from published feeding studies. The nine amino acid cells in section 2 come from somewhere else entirely: raw USDA composition, divided by requirements read out of the DRI report. Two routes that share no data. At the default basket they land on the same answer — lysine is the lowest cell of the nine, at roughly half your requirement, while the other eight run from 79% to 123%. Nobody told the arithmetic what to find. If the two had disagreed, this section would say so instead, because a page that only reports its confirmations is not measuring anything.
Read the whiskers, not the bars. Those are one
standard deviation, and for egg (±58) and whey (±25) they are enormous — the review's own
conclusion is that variation within a protein type, driven by processing and genotype,
is large. A single point value for "egg" or "whey" would be the misleading thing here, so
the spread is drawn. Two further cautions we have to state against ourselves: these are scored
against the 0.5–3 year reference pattern, and a different pattern moves every score; and
whey scoring 85 with histidine limiting will surprise anyone expecting whey above 100 — that is
the pattern choice plus real between-study variance, not an error.
| hemp preparation | PDCAAS | DIAAS | true digestibility | what it actually is |
|---|---|---|---|---|
| defatted hemp hearts | 44.0% | 0.45 | 90.4% | this is the row that maps to the hearts in your basket — dehulled hemp seed |
| hemp protein concentrate 1 | 42.3% | 0.43 | 91.9% | mechanical air-classification, 66.2% protein |
| hemp protein concentrate 2 | 43.9% | 0.45 | 87.0% | mechanical air-classification, 53.6% protein |
| casein — the reference | 100% | 1.03 | 96.3% | what a score of 1.00 means |
| ↳ Nosworthy et al. 2023, Food Science & Nutrition. Rat bioassay, and that limit ships with the number — rodent digestibility is not human ileal digestibility, and DIAAS is properly defined on the latter. One dissenting paper names tryptophan rather than lysine as limiting, but it studied a single isolate and reported leucine and isoleucine as one combined row, which is not a standard pairing. Two independent sources say lysine; it is named here as lysine with the exception recorded. | ||||
The consequence for the ledger above, stated plainly:
at a PDCAAS near 0.44, roughly 44% of hemp's protein nitrogen is usable for synthesis in a
single-food idealisation, and the rest is deaminated. So the protein row overstates usable
protein by close to a factor of two — unless the day's other foods supply the lysine, which is
precisely what the brake above says they can. Both halves are true and neither is allowed to
travel without the other.
6 · FIBRE — the outcome evidence, carrying the certainty its own authors assigned
A correction this page owes on its own hero food first. Hemp's fibre is
overwhelmingly in the hull — hulls run 77–84% dietary fibre, whole seed 28–36%. Hearts
are the dehulled fraction, and measure 4.0 g per 100 g. Whole-seed fibre figures must never be
quoted for hearts, and this page uses the hearts number. Hemp hearts are an excellent protein and
fat source and a poor fibre source; saying so costs nothing but a slogan.
Below is the largest synthesis in the field —
Reynolds & Mann et al., The Lancet 2019: 185 prospective publications, just under
135 million person-years, plus 58 randomised trials, commissioned by WHO. The certainty
column is the authors' own GRADE rating, not ours, and it is the most important column.
| outcome | design | effect | the authors' GRADE |
|---|---|---|---|
| body weight | randomised trials | 0.37 kg lower | HIGH |
| total cholesterol | randomised trials | 0.15 mmol/L lower | MODERATE |
| systolic blood pressure | randomised trials | 1.27 mmHg lower | MODERATE |
| HbA1c | randomised trials | not significant | LOW |
| all-cause mortality | observational | RR 0.85 | MODERATE |
| coronary heart disease incidence | observational | RR 0.76 | MODERATE |
| type 2 diabetes incidence | observational | RR 0.84 | MODERATE |
| colorectal cancer incidence | observational | RR 0.84 | MODERATE |
| stroke mortality | observational | RR 0.80, CI crosses 1 | VERY LOW |
THE TENSION YOU MUST SEE, and it is the whole reason this
table is drawn this way. Look down the two columns together. Every disease endpoint is
observational — no trial has ever randomised people to high or low fibre and counted deaths —
so the 15–30% mortality reductions are associations, open to the obvious confound that people who
eat more fibre differ in smoking, activity, income and whole diet. And every outcome with High
or Moderate certainty has a small effect: 0.37 kg of body weight, 1.27 mmHg of blood pressure
whose confidence interval barely clears zero. The outcomes with big effects have the weakest
evidence; the outcomes with the strongest evidence have small effects. A reader who takes
"HIGH certainty" to mean "large benefit" has been misled, and this page will not be the thing that
misleads them. The authors themselves write only that the relationships "may be" causal.
Two limits that bear directly on this page's own
interests. First: the review states plainly that "limited data were available regarding
specific sources or subcategories of dietary fibre" — the signal is dominated by cereal
fibre. Hemp fibre cannot inherit these results, and this page does not let it. Second:
the trials excluded powdered fibre supplements by design, and the authors note supplemental-fibre
reviews show "only small effects." This evidence supports fibre-rich foods, not isolated
fibre — which cuts against every product this project might one day want to sell. Also
recorded: the authors name one harm, that high intakes may be deleterious where iron or mineral
status is already borderline. And the optimum they identify is 25–29 g/day, against typical
population intakes below 20.
What is mechanism and what is outcome. That colonic
fermentation yields acetate, propionate and butyrate, and that butyrate is the primary fuel of the
colonocyte, is established biochemistry. That any specific clinical outcome in the table
above is mediated by those short-chain fatty acids in humans is a plausible and
much-studied hypothesis, not a demonstrated causal chain. Mechanism and outcome are joined
here by inference. That seam is where this page stops and the graded evidence
deck begins.
7 · FAT, DISAGGREGATED — "saturated fat" is never one number here
That category collapses two things a body routes
differently — MCTs (C6–C12) leave the gut by the portal vein straight to the liver, LCTs
(C14+) by chylomicron through the lymphatics. USDA publishes fatty acids individually by chain
length, so this split is not an interpretation we add; it is a column most labels stop
showing. Lauric acid (C12) gets its own row because it is an MCT by chain length yet is
absorbed substantially like an LCT — and coconut oil, this basket's own MCT source, is roughly
half lauric. Publishing that against our own favourite fat is the point: a page that only
disaggregates other people's food is marketing.
8 · THE ARBITRAGE — what the measurements say people actually miss
THE BRAKE, and it ships with the number every time.
"Below the EAR" is a statistic about dietary intake, not a deficiency rate. Measured by
serum 25(OH)D, about 5% of people are at risk of vitamin D deficiency against 96%
below the dietary EAR — because the board set that RDA "on the assumption that people receive
minimal sun exposure." Vitamin E intake is under-captured because cooking fats go uncounted,
and choline has no EAR at all. Publishing 96 / 88 / 74% without this paragraph would be
technically accurate and substantively misleading, which is the failure mode this page exists to
refuse.
And the half that never gets quoted: selenium,
niacin, riboflavin, phosphorus, iron (6%), thiamin, B12 and protein are not
population shortfalls. A page that shows only the alarming half is a campaign, not a measurement.
9 · THE QUEST — a plan for a day, a week, a month, a year
Each rung is a real span of time and a different strength of evidence.
You cannot claim a year-scale result from a day of remembering what you ate, so the rungs do not
let you. Every quest is built and checked in Beehive University.
…AND HOW ANY OF IT COULD BE BELIEVED
Four different questions hide inside "how do we know it worked", and they have
four different answers. Collapsing them is how wellness claims get made.
| the question | the honest answer here | who can check it |
|---|---|---|
| documented | The plan is a file you compose and hold. This page has no account, no server and no database — it computes in your browser and stores nothing. The hive never receives your food log. | You, completely. There is no custodian to trust, because there is no custodian. |
| monitored | By nobody, by design. An operator who watches your intake is a party who can be subpoenaed, breached or sold. Autonomy means every factor is self-verifying — so monitoring is something you do to your own record, not something done to you. | You. This is a deliberate capability we refuse to build, not one we have not built yet. |
| measured | Two things get measured and they are not the same thing: what the food contains (this page, from published composition), and what your body did with it (a blood panel). Only the second is a measurement of you. | Anyone, for the first. An accredited lab under chain of custody, for the second. |
| independently verified |
This is where most of this page is genuinely strong, and where self-report is genuinely worthless. Every requirement traces to a published NASEM value; every food figure to a USDA record ID you can refetch; the EER equation is printed in full. A stranger can recompute this entire page from primary sources and catch us if we are wrong. What no stranger can verify is that you actually ate it. | Anyone with a browser, for the arithmetic. Nobody, for adherence — unless you choose to attach evidence, which is what the ladder below is for. |
THE EVIDENCE LADDER — every claim carries the tier it
was earned at, and never a higher one.
| tier | what it evidences | what it cannot support |
|---|---|---|
| 0 · composed | a plan exists and its arithmetic is correct | anything at all about a body. A plan is a hypothesis. |
| 1 · self-reported | your own account of what you ate | the weakest tier in nutrition science and it is not close. Dietary self-report is systematically under-reported, and the error is not random — it varies with body weight, which biases exactly the comparison a page like this invites. Never present a tier-1 result as a finding. |
| 2 · receipt-backed | that the food entered your household — purchase records, a scale reading, a lot number on a bag | that you ingested it, or when. It closes the honesty gap on supply, not on intake. |
| 3 · assayed | what is measurably in your blood — serum 25(OH)D, ferritin, B12, a lipid panel, omega-3 index. The only tier that measures the body rather than the food. | attribution. A changed biomarker does not name its cause; that needs a design, a control and a pre-registered question — which is the BiGen lane, off this page. |
The one number on this page that already sits at tier 3
for everyone: the vitamin D brake in section 6. Serum measurement says roughly 5% are deficient
where dietary intake says 96% fall short. That gap between what people eat and what is in their
blood is the reason the ladder exists, and it is why this project will not publish an
intake statistic as a health finding.
10 · SOURCES — cite, or the cell stays silent
External citations open in a new tab so this page and session stay put.