How Probiotic Potency Is Measured — CFU, AFU, and What Survives
Two probiotic capsules can both say "10 billion" on the label and still be measuring completely different things. Here is how the counting actually works — with the batch certificate, production process and gastric-tolerance data behind our own formula shown in full.
The short version
- CFU counts only the cells that can grow into colonies on a plate. It says nothing about cells that are alive but not dividing.
- AFU/TFU uses flow cytometry to count all intact microbial cells, live or damaged. The international method is ISO 19344.
- For Akkermansia muciniphila — a strict anaerobe that does not divide easily on plates — total-cell counting is the more honest unit.
- A batch certificate should disclose the counting method, the specification and the tested result. The supplier-reported figures are summarized below, although the original certificate is not reproduced here.
Why "the same number" can mean different things
Pick up two probiotic capsules from any shelf and compare labels. One says "50 billion CFU". Another says "10 billion AFU". A third just says "probiotic blend". All three are trying to answer the same question — how many bacterial cells are in here — and all three are answering it with a different measuring stick.
The confusion is not marketing spin, at least not originally. It comes from a genuine measurement problem: there is more than one way to define whether a bacterial cell "counts", and the method you choose changes the number by an order of magnitude. If you take a digestive support product seriously enough to compare labels, it pays to know which stick is being used.
CFU: the classic plate count
CFU stands for colony-forming unit. The method is exactly what it sounds like: you plate a diluted sample on nutrient medium, incubate it, and count the visible colonies that grow. Each colony is one "unit" [1].
CFU has one big virtue — it only counts cells that can reproduce, which is a reasonable proxy for "alive" — and several blind spots. Cells that are alive but injured, dormant, or simply slow to divide under the test conditions will never form a colony, so they never get counted. And some organisms divide in chains or clusters, so one colony may trace back to more than one cell. The international probiotic definition agreed by ISAPP — "live microorganisms that, when administered in adequate amounts, confer a health benefit on the host" — is deliberately built on the idea of live cells, which is why CFU became the industry's default unit [1].
AFU and TFU: counting with flow cytometry
The second approach does not grow anything. Flow cytometry pushes cells single-file past a laser and counts every intact cell it detects, live or damaged. Results are reported as AFU (active fluorescent units) or TFU (total fluorescent units), and the reference method is the ISO 19344 standard [2].
For most bacteria, AFU/TFU numbers come out higher than CFU numbers for the same sample, because the cytometer sees cells that the plate cannot. Neither number is "wrong". They answer different questions: CFU asks "what can replicate on this plate?", flow cytometry asks "what intact cells are physically present?"
Some bacteria are uniquely badly served by plate counts
Akkermansia muciniphila is a strict anaerobe, first isolated in 2004 from a human intestinal sample, that grows slowly and only under oxygen-free conditions [3]. Plate-counting it is genuinely hard, and under-counting is the expected failure mode. That is why serious Akkermansia products — and the regulatory filings behind them — specify the dose in TFU (total fluorescent units) rather than CFU. The FDA filing for our strain, for example, states its conditions in TFU per day [7].
What a real batch certificate looks like
Anyone can write "quality tested" on a website. The document that actually carries information is the certificate of analysis for a specific production batch. The supplier's technical dossier reports the following figures for a pasteurized Akkermansia powder batch (SH24090222, tested 1 October 2024). The original certificate is not reproduced on this page; the figures should be verified against the manufacturer's documentation.
Four details in that certificate are worth reading closely:
| Line item | What it tells you |
|---|---|
| TFU (cells)/g ≥2.0×10¹¹, tested 5.0×10¹¹ | The potency claim is a total-cell count measured by flow cytometry, with a tested result above the specification — not a number chosen for the label. |
| Live bacterial count <10 CFU/g | This batch is deliberately inactivated. A live-cell count near zero is the specification working, not a defect. The form (live vs. pasteurized) is a separate decision from the counting method. |
| E. coli <10; mold ≤50 | Contaminant limits sit alongside the potency lines. A potency number without contaminant lines is half a certificate. |
| Shelf life 24 months, room temperature, double-layer PE + aluminum foil packaging | Potency is a claim about the end of storage, not the day of filling. Packaging and storage conditions are part of the potency claim. |
Not all bacteria are counted the same way
The counting method interacts with the biology of each organism. A few examples show why:
- Spore formers. In a spore based probiotic such as Bacillus coagulans — often searched as a bacillus coagulans probiotic — the product is the dormant spore, which survives gastric acid well and germinates in the intestine. Viability testing is straightforward, because the spore germinates reliably on the plate.
- Yeasts. A s boulardii probiotic uses Saccharomyces boulardii, a yeast rather than a bacterium, so it is naturally resistant to antibacterial antibiotics — a relevant trait the counting methods do not capture either way.
- Vegetative lactobacilli. L plantarum probiotics deliver actively dividing cells that are more fragile in storage and in stomach acid, which is why quality products pair them with protective delivery formats and test potency at end of shelf life.
- Strict anaerobes like Akkermansia. As above: plate counts under-count, so total-cell flow cytometry is the honest unit.
Reading a label without being misled
Five checks separate a serious label from a decorated one:
- Which unit, and at what time point? "50 billion at time of manufacture" and "50 billion at expiry" are different promises. Potency at expiry is the meaningful one.
- Is the counting method stated? For Akkermansia, look for TFU. For typical lactobacilli and symbiotic probiotics (probiotic-plus-prebiotic combinations), CFU at expiry is standard and fine.
- Is the strain identified? A name without a strain number ("Lactobacillus plantarum") is a species, not a strain. Strain-level identity is where the science lives.
- Are storage conditions consistent with the claim? A room-temperature claim should come with packaging and stability data behind it, like the 24-month specification in the certificate above.
- Is the full formula disclosed? Serious digestive nutritional supplements — and any probiotic for gut health you plan to take daily — should disclose every ingredient and every milligram. Proprietary blends hide both dose and ratio.
What about surviving stomach acid?
Counting answers "how many are in the capsule". Gastric tolerance answers "how many arrive". These are separate tests. For the Akkermansia strain in our formula, the manufacturer's safety and characterization study — published in the peer-reviewed journal Foods in 2024 — reported good gastrointestinal tolerance in vitro, with roughly 84% and 82% survival across simulated gastric and intestinal passage assays [4]. That number describes the strain under test conditions; it is not a claim about results in any individual person, and it is one of several reasons the formula uses a protective delivery matrix rather than counting on unprotected cells to survive the trip.
What these numbers cannot tell you
Cell counts, tolerance assays and batch certificates describe the product. They do not predict how any individual will respond, and they are not a measure of clinical benefit — that comes from intervention trials in people, which for Akkermansia remain small and early [5][6]. If a label implies that a big number equals a big outcome, the number is doing work the science has not done.
Where this fits in a real routine
Our GLP-X +MAX delivers Akkermansia muciniphila AKK PROBIO in a protective time-release matrix, with the batch-level TFU specification, counting method and 24-month stability data shown above. Every milligram of the ten-ingredient formula is disclosed on the label.*
See GLP-X +MAX →Common questions
Are AFU bigger numbers than CFU, and does that mean more?
AFU/TFU counts every intact cell, so the number is usually higher than a CFU count of the same sample. Neither is inflated or deflated — they measure different things, and the counting method should be stated on the label.
Why is the live count under 10 CFU/g in the certificate if this is a probiotic?
Because that batch is the pasteurized (inactivated) powder — the spec requires live organisms to be absent, and the test confirms it. The live vs. inactivated form is a separate product decision; a live-form product requires its own appropriate viability documentation; this pasteurized-batch certificate does not establish its CFU potency.
Does "50 billion CFU" guarantee results?
No. Cell count describes the dose, not the outcome. Clinical benefit has to be shown in human trials of the specific strain at the specific dose — a number on a front label does not substitute for that evidence.
What is a symbiotic?
A product combining live microorganisms with a substrate (prebiotic) selectively used by them. It is a formulation concept, not a potency unit — the counting rules above still apply to its probiotic component.
How long do probiotics stay potent in the bottle?
It depends on the strain, the form and the packaging. The certificate above specifies 24 months at room temperature in double-layer PE and aluminum foil packaging for the inactivated powder. Live-cell products are typically more fragile and may need refrigeration — check the label's own stability data, not the category average.
References
- Hill C, Guarner F, Reid G, et al. Expert consensus document: The International Scientific Association for Probiotics and Prebiotics consensus statement on the scope and appropriate use of the term probiotic. Nat Rev Gastroenterol Hepatol. 2014;11(8):506–514. doi:10.1038/nrgastro.2014.66. PMID 24912386.
- ISO 19344:2015. Milk and milk products — Starter cultures, probiotics and fermented products — Quantification of lactic acid bacteria by flow cytometry. International Organization for Standardization. iso.org/standard/63421.html.
- Derrien M, Vaughan EE, Plugge CM, de Vos WM. Akkermansia muciniphila gen. nov., sp. nov., a human intestinal mucin-degrading bacterium. Int J Syst Evol Microbiol. 2004;54(Pt 5):1469–1476. doi:10.1099/ijs.0.02873-0. PMID 15388697.
- Ma X, Tian M, Yu X, Liu M, Li B, Ren D, Wang W. Characterization and Preliminary Safety Evaluation of Akkermansia muciniphila PROBIO. Foods. 2024;13(3):442. doi:10.3390/foods13030442. PMID 38338577.
- Depommier C, Everard A, Druart C, et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nat Med. 2019;25(7):1096–1103. doi:10.1038/s41591-019-0495-2. PMID 31263284.
- Cani PD, Depommier C, Derrien M, Everard A, de Vos WM. Akkermansia muciniphila: paradigm for next-generation beneficial microorganisms. Nat Rev Gastroenterol Hepatol. 2022;19(10):625–637. doi:10.1038/s41575-022-00631-9.
- FDA (R. Philip Yeager, Director, Division of Research and Evaluation). Response to New Dietary Ingredient Notification 1468 — Akkermansia muciniphila AKK PROBIO. May 20, 2026. Letter on file; conditions of use stated in TFU per day.
*These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
This article is educational and is not medical advice. It is not intended to diagnose, treat, cure or prevent any disease, and it is not a substitute for consultation with a qualified healthcare professional.
Batch-level manufacturing data (certificate, process flow, tolerance assays) are reproduced from the strain manufacturer's technical dossier and cited where published. Where a study was conducted in vitro or in animals, that is stated in the text.