Label literacy

CFU Or Milligrams? How To Read A Probiotic Label

Reviewed by Dr. Susan Vogler, DO, FACOG, MSCP

Two units, and why only one of them means anything

Open any serious probiotic and the front of the bottle carries a number followed by the letters CFU, usually in billions. Open a great many supplements with a probiotic blend in them and you find a milligram figure instead. Those two numbers are not different ways of saying the same thing, and the difference is the single most useful piece of label literacy in this category.

A milligram figure describes mass. If a manufacturer puts 50 mg of freeze-dried powder into a capsule, the panel can honestly say 50 mg, and that statement is compatible with almost any number of living organisms including very few.

A colony forming unit figure describes how many of them are alive. It is the only number that corresponds to what the research measures, and NCCIH's consumer page on probiotics is explicit that effects depend on both which strain and how much.

What a colony forming unit actually counts

The name is literal. A sample is diluted and spread on a growth medium, and every organism capable of reproducing forms a visible colony. Count the colonies, multiply by the dilution, and you have the number of viable organisms in the original sample.

It is called a colony forming unit rather than a cell because a clump of cells that grows together forms one colony and counts once. It is a slight undercount by design, and it is the number that matters, because an organism that cannot multiply cannot establish itself anywhere.

Two things follow. Dead cells do not count, which is the whole point. And the count is a measurement rather than a calculation, so it has to be performed rather than derived from a formulation sheet.

Why milligrams cannot be converted into a count

People sometimes ask for the conversion factor. There is not one, and the reasons are worth listing because each is a separate gap.

• Species and strain differ. Cells differ in size and in how well they survive drying, so the same mass of two powders holds different numbers.
• The carrier is part of the mass. Freeze-dried cultures are almost always blended with maltodextrin, rice starch or a similar bulking agent. The ratio is a formulation decision and it is never printed.
• Dead cells weigh the same as live ones. A powder that has lost most of its viability weighs exactly what it did when it was made.
• Manufacturing conditions vary. Heat and moisture during tabletting or encapsulation kill organisms, and the amount lost depends on the process.

Put those together and a milligram figure for a live-organism blend is not a weak version of a colony count. It is a different quantity that cannot be turned into one.

At manufacture, or at expiry? The number that moves

Even a label that prints a count can print two very different numbers, and the difference is a real one.

A count at time of manufacture is the highest number the product will ever have. From that moment viability falls, slowly in a cool dry cupboard and rather faster in a warm one. By the time a bottle has crossed a country in summer, sat in a warehouse and then in a bathroom, the count can be a fraction of what it was.

A count at the end of shelf life is a guarantee that the stated number is still present on the last day of the best-before period. It requires the manufacturer to overfill at production and to have stability data, and it is the number a buyer actually receives.

When a label prints a count without saying which it is, assume manufacture. When a label prints no count at all, neither question arises, which is convenient for everybody except the reader.

Three lines of print, three very different levels of disclosure, and no difference in cost to the manufacturer.

What the Femicore label prints

The third row. The panel carries a Probiotic Blend at 50 mg, names five strains inside it with their designations, and prints no colony count anywhere on the bottle, for the blend or for any single strain. Neither does the sales page, and neither does the material this retail desk holds.

To be fair about what it does do: naming five strains with designations rather than writing “probiotic blend” is a genuine disclosure that many competitors skip, and it lets a reader look up the research on L. crispatus LCr86 specifically. It is half of the necessary information and it is the harder half to fake.

What it does not do is let anyone compare this product with a trial, or with a competitor that prints its count. That is not an accusation of anything; it is a description of what can be known from the bottle, and it is why the label page lists the colony count first among the figures this panel omits.

How many organisms do the trials actually use?

The honest answer is that it depends on the strain and the outcome, which is exactly why the count has to be printed rather than assumed.

Trials in this field state their dose as a matter of course. The phase 2 trial of L. crispatus in recurrent urinary tract infection states the count per application. The oral colonisation trial from 2003 states a daily dose for each of its two strains. The 2024 pregnancy trial does the same. Nobody in research treats the number as optional.

Quoting a figure from one of those trials as though it applied to a different strain would be exactly the error this article is about, so this page does not do it. The useful takeaway is structural: every study you might want to compare a product against reports a count, so a product that does not report one cannot be compared with any of them.

A five-point test for any probiotic label

1. Is there a colony forming unit figure at all? If not, nothing else on the panel can be evaluated.
2. Is it per strain or for the blend? A blend total across five strains leaves the split unknown, exactly as a milligram blend does.
3. At manufacture or at end of shelf life? If unstated, assume manufacture.
4. Are the strains named with designations? Species alone is not enough, because research is done on strains.
5. Is there a storage instruction? A product that says nothing about heat is either very stable or has not thought about it, and the label rarely says which.

Run that on two products side by side and the comparison usually takes under a minute. It is the same discipline the nine-point checklist applies across the whole category, and that page scores this website’s own product against it including the rows it fails.

Why a manufacturer might not print the number

Three reasons, and only one of them is sinister.

The formula is a trade secret. A proprietary blend exists to stop competitors reading a formulation off a label, and that is a legitimate commercial interest even when it is inconvenient for the buyer.

Stability data is expensive. Guaranteeing a count at end of shelf life requires testing across the storage period and overfilling to cover the decline. A smaller operation may genuinely not have that data rather than be hiding it.

The count is not impressive. This is the one that matters to a buyer, and from outside the factory it is indistinguishable from the other two. That is the whole problem with an absent number: silence looks identical whatever the reason for it.

Which is why the right response is neither outrage nor a shrug. It is to note the gap, weigh it against what the label does disclose, and price the uncertainty into the decision. For this product that means a five-strain blend at an unknown count alongside a cranberry extract that is standardised, which is the trade set out on the ingredients page.