Why Proprietary Blend Doses Matter

FDA's rule for proprietary blends takes up a few lines of federal regulation. A blend on a Supplement Facts panel must:

• carry a name, either "Proprietary Blend" or another descriptive or made-up name the company picks;
• show the total weight of the blend's ingredients on the same line as that name;
• list those ingredients indented beneath it, from most to least by weight;
• list separately, with their own amounts and %DV, any ingredients that have a Daily Value, such as vitamins and minerals;
• mark the total with the symbol for "Daily Value not established."

Notice what the list leaves out: any limit on how many ingredients a blend can hold, or on how little of any one it can contain.

A proprietary blend tells you which ingredients are in it and which one weighs the most. It hides how much of each there is, and with that, the ratio between them.

Companies usually say blends protect a formula from copycats. That's a fair worry for anyone who has spent years on a formula. My view is that a competitor with a lab can often measure a product anyway, so the person a blend keeps in the dark most reliably is the buyer.

Classical Chinese formulas were written down with an amount for each herb. Chinese medicine is the tradition behind the Five Phases, also called the Five Elements, and one of its best-known formula books, the Shang Han Lun, was compiled by Zhang Zhongjing around 200 CE. It gives Gui Zhi Tang as five ingredients: gui zhi (cassia twig), 3 liang; shao yao (peony root), 3 liang; gan cao (licorice root, roasted), 2 liang; sheng jiang (fresh ginger), 3 liang; and da zao (jujube), 12 pieces.

The detail I love comes later in the same book. The same five ingredients appear again with the peony doubled to 6 liang, under a new name: Gui Zhi Jia Shao Yao Tang, "Gui Zhi Tang with added peony." Change the ratio, and the book counts it as a different formula. Scholars still argue about how much a Han-dynasty liang weighed, so neither recipe converts neatly to milligrams. Both still give you the proportions, which is more than a proprietary blend will.

You can't get exact amounts from a proprietary blend, but the order gives you a ceiling for each ingredient. The rule I use: the ingredient in position N weighs at most the blend's total divided by N. Every ingredient above it weighs at least as much, so N ingredients at that weight would already fill the blend. The second ingredient is at most half the total, and the third at most a third. Only the first ingredient has a floor. It weighs at least the total divided by the number of ingredients.

Here's the math on a made-up 600 mg blend with six ingredients:

The right-hand column is the uncomfortable one. From the second ingredient down, any of them could be a trace.

One wrinkle changes the math. The blend's total covers only the ingredients without a Daily Value. A vitamin or mineral sitting inside the blend is listed with its own amount and isn't counted in that total, so leave it out when you divide. And a heading on its own hides nothing: if every ingredient under a blend's name shows its own amount, you already have the numbers the ceiling test estimates.

Comparing a blend with the research takes five steps, and the fourth can rule a product out with simple division.

1. 1Find the dose used in human studies. NIH's Office of Dietary Supplements (ODS) publishes fact sheets on many vitamins, minerals and herbs that describe the human research, including the amounts studied. For the studies themselves, PubMed is free, and the methods section of each paper gives the daily dose and the form.
2. 2Match the form. Research on a root extract standardized to a certain compound tells you little about a powdered leaf. Check the plant part, the extract ratio and any standardization (How to Read a Supplement Label shows where each one appears).
3. 3Match the daily amount. Multiply the amount per serving by the number of servings the label recommends per day, then set that against the daily dose in the study.
4. 4Run the ceiling test. Say the studies used 300 mg a day, and the ingredient sits fourth in a 600 mg blend taken once a day. It can't be more than 150 mg, so at that serving the product can't reach the studied dose.
5. 5Keep ingredient research and product research apart. A study of one ingredient says little about a product that combines it with five others. A formula has been tested as a formula only if someone ran a trial on it.

To line up labels side by side, NIH's Dietary Supplement Label Database (DSLD) holds more than 200,000 US supplement labels, with each ingredient's name, form and amount as printed. It's free to search, and an easy place to lose an evening.

When you ask a company about its proprietary blend, start with the amount of each ingredient per serving. Some companies will share it by email even when the label doesn't. Then ask whether each herb is an extract or a whole powder, and which part of the plant was used. For herbs, those two answers can matter as much as the dose. If the company won't say, you're left with the ingredient order and the total.

If there's a bottle with a blend in your cupboard right now, try the ceiling test on its fourth ingredient. It takes about a minute, and it shows you the whole problem in one number.