Labels
Reading a proprietary blend
A proprietary blend prints two things: a total weight, and an order. It prints no individual amounts, and it is not required to. What almost nobody notices is that those two things together force a hard ceiling on every ingredient in the blend — a real number, derived from the label, that the label does not contain. This post works one out, six times, on the bottle this site is about.
What a blend actually prints
Start with the row in front of us. This panel has three lines. Two are ordinary nutrients with weights and percentages: Vitamin B3 (as Niacin) 20 mg, 125% and Zinc (as Zinc Oxide) 11 mg, 100%. The third reads Proprietary Blend — 570 mg, and underneath it, indented, six ingredient names with nothing beside them.
Add the panel up and a capsule declares 601 mg of material. The blend is 570 of it — about 95% by weight. So the two rows a buyer can check are the two smallest. That is the situation, and it is entirely lawful.
What the blend row gives you is therefore exactly two pieces of quantitative information: 570 mg, and the sequence in which the six names appear. Most readers treat the sequence as decoration — a list has to be in some order, after all. It is not decoration. It is the more useful of the two.
The rule, quoted in full
The sequence is not arbitrary because federal labelling guidance says it may not be. Here is the paragraph, from the FDA’s Dietary Supplement Labeling Guide, reproduced without cuts:
FDA — Dietary Supplement Labeling Guide, Chapter IV
“You must identify proprietary blends by use of the term “Proprietary Blend” or an appropriately descriptive term or fanciful name. On the same line, you must list the total weight of all “other dietary ingredients” contained in the blend. Indented underneath the name of the blend, you must list the “other dietary ingredients” in the blend, either in a column or linear fashion, in descending order of predominance by weight.”
Three requirements. Name the blend. Print its total weight on the same line. List the constituents underneath in descending order of predominance by weight. And no requirement at all to print an individual weight for any of them — which is why the niacin and the zinc on this panel carry figures and the six blend ingredients do not. Ingredients with an established Daily Value must be itemised; the rest need not be.
So this label is not evading a rule. It is following one, and the rule is what permits 95% of a formula to go unquantified. Any useful reading of a blend starts from that rather than from indignation, because indignation does not produce numbers and the descending-order clause does.
The proof, in one sentence
Descending order means the first-listed ingredient weighs at least as much as the second, the second at least as much as the third, and so on down to the sixth. From that, plus the total, an upper bound follows for every position. The argument is short enough to check in your head:
The whole of the reasoning
Suppose the ingredient in position n weighed more than the total divided by n. Every ingredient above it weighs at least as much as it does, so those n ingredients together would weigh more than n times the total divided by n — that is, more than the total. Impossible. Therefore the ingredient in position n cannot exceed the total divided by n.
That is the entire method: divide the blend total by the position number. It needs no chemistry, no laboratory and no information beyond the two things every compliant blend must print. It works on a blend of three and a blend of thirty, in a shop aisle, on a telephone.
It also has a useful shape. The bound on the first ingredient is the total itself, which tells you nothing, and it tightens with every step down the list. In a blend, the interesting arithmetic is at the bottom.
The worked example: 570 mg across six
Six divisions. The total is 570 mg and the positions run one to six, in the order the label prints them.
| Position | Ingredient, as printed | The division | Ceiling |
|---|---|---|---|
| 1 | L-Citrulline | 570 ÷ 1 | 570 mg |
| 2 | L-Carnitine | 570 ÷ 2 | 285 mg |
| 3 | Pine (Pinus pinaster) Bark Extract | 570 ÷ 3 | 190 mg |
| 4 | Maca (Lepidium meyenii) Root Extract | 570 ÷ 4 | 142.5 mg |
| 5 | Grape (Vitis vinifera) Skin Extract | 570 ÷ 5 | 114 mg |
| 6 | Saffron (Crocus sativus) Stigmas Extract | 570 ÷ 6 | 95 mg |
Six numbers, none of them printed on the bottle and all six forced by it. Two things about the table are worth pausing on. The first row is no use: bounding the largest ingredient by the whole blend is arithmetic that tells you nothing. And the six ceilings cannot all be attained at once — six ingredients at their individual ceilings would total 1,396.5 mg, well over twice the blend. Each bound is true separately; together they are far looser than reality.
Where the ceiling starts to bite
A ceiling only becomes interesting when there is a published figure to set it against. On this blend there is exactly one, and it belongs to the ingredient in position two.
The NIH Office of Dietary Supplements records athletic-performance studies of L-carnitine using 1 to 4 grams a day, and summarises the results as mixed. The ceiling in position two is 285 mg. So whatever is actually in this capsule, it is at most a little over a quarter of the lowest amount anybody has studied — and quite possibly a great deal less, since a ceiling is not an amount.
Notice what that argument did not require: knowing the real figure. The bound alone settles it, because the bound is already below the range that matters. That is when the method pays.
For the other five ingredients the method returns a number and nothing to compare it against. L-citrulline, pine bark, maca and saffron return no monograph at the NIH Office of Dietary Supplements or NCCIH under the names printed here. Grape skin has none either; the monograph that exists is about the seed, a different part of the plant. So five ceilings sit there, correctly computed and unusable, which is itself worth knowing.
What a ceiling is not
This is where people who have just learned the trick go wrong, and the error is worth naming because it is made in both directions.
A ceiling is not an amount. “At most 285 mg” is a true statement about this label; “285 mg” is an invented one. The blend could be L-citrulline at 565 mg and the other five at a milligram apiece, or six at 95 mg exactly, and the printed label is identical. Anyone quoting a per-ingredient figure for a blend is reading a document you have not seen, or making it up.
A ceiling is not a typical value either. There is no distribution here to be typical of. The saffron in position six cannot exceed 95 mg and may be one milligram, and nothing on the bottle distinguishes those two worlds.
And the asymmetry is total. Descending order gives an upper bound and never a lower one. You can demonstrate that an ingredient is not large; you can never demonstrate, from a blend alone, that it is not vanishing. That is what a proprietary blend costs a reader.
Doing this on any bottle, in about a minute
The whole method, portable, with nothing specific to this product in it.
Find the blend line and read the total
It is on the same line as the blend’s name, and it must be there. If there is no total, the label is not compliant and the rest of this is moot.
Count the constituents, and count the position you care about
Starting at one, in the order printed. Do not reorder them, do not alphabetise them, and do not assume the interesting one is first.
Divide
Total divided by position. That is the ceiling for that ingredient, in whatever units the total is printed in.
Divide again for the average
Total divided by the number of constituents. It says nothing about any single ingredient and a good deal about the order of magnitude the blend works in.
Set the ceiling against a published dose
The NIH Office of Dietary Supplements and NCCIH are where to look. If the ceiling is already below the studied range, you have your answer without ever learning the amount.
Run it on this bottle and the summary is: 570 of 601 mg, about 95% of the capsule by weight, knowable only as six ceilings, of which one is decisive and five have nothing to be measured against. That was reached with a printed total, a printed order and division — and it is more than the front of the box will ever tell you.
One last use for the method. If a manufacturer ever does publish a breakdown, the ceilings are how you check it: every figure must fall under its bound and the six must sum to the printed total.
Read the blend row first, then decide
Thirty capsules, one a day. Two named nutrients with percentages that both check out, and then 570 of the 601 milligrams in the capsule inside a proprietary blend with no amounts on it. The live price and the 60-day return terms are both on the seller’s checkout screen.
60-day window · Questions first? Call +1 (323) 348-1373