Two probiotic products can carry identical numbers on the front of the pack and deliver quite different quantities of living bacteria to the person taking them. Neither manufacturer has to lie for this to happen. The difference lies in a detail that most labels do not make prominent: when the count was measured.
A colony forming unit count is a snapshot of how many cells were alive and capable of dividing at the moment of testing. Probiotic bacteria die steadily in storage, at a rate governed by temperature, moisture and the strain itself. A count taken as the powder leaves the blender and a count taken at the end of the stated shelf life can differ by a large factor, and both are honest measurements of different things.
Key takeaways
- A colony forming unit counts a clump that grew into a colony, not an individual bacterial cell.
- Counts at manufacture and counts at expiry describe different things, and only the second is what you consume.
- Plate counting carries substantial inherent uncertainty, so a count is an estimate with a range rather than an exact figure.
- Species labelling without a strain designation is not enough, since effects documented for one strain do not transfer to others.
- Moisture and temperature drive viability loss more than time alone, which makes storage conditions part of the claim.
What a Colony Forming Unit Measures
The name is precise and is usually read carelessly. A colony forming unit is whatever grew into one visible colony on a plate. It is not a cell count.
The method is old and mechanically simple. A sample is diluted in stages, a measured volume of a suitable dilution is spread on nutrient agar or mixed into it, and the plate is incubated under conditions the target organism can grow in. Each viable unit that lands on the plate divides repeatedly in place until it forms a mass of cells visible to the eye. The colonies are counted, multiplied by the dilution factor, and reported.
The gap between the count and the number of cells arises because bacteria do not arrive singly. Many probiotic species grow in chains or pairs, and freeze-dried powders contain aggregates held together by the drying matrix. A chain of several cells that lands together produces one colony and is counted as one unit. The count therefore systematically understates cell number, by a degree that depends on the organism and the formulation and that is rarely quantified.
The word viable carries the other half of the definition. Only cells that can divide under the plating conditions are counted, which excludes cells that are alive but injured, and cells in a state where they are metabolically active but will not grow on a plate. Freeze-drying and passage through a product’s shelf life both produce sublethally injured cells, which may recover in a gentler environment such as the gut but which count as nothing on a plate. Laboratories address this partially by including a resuscitation step in a non-selective medium before plating, but the effect is not fully eliminated.
Alternative methods sidestep some of this. Flow cytometry with fluorescent stains that distinguish intact from damaged membranes counts individual cells rather than colonies, gives an answer in hours rather than days, and includes cells that would not grow on a plate. Its results are reported in different units precisely because they are not equivalent, and they are typically higher than plate counts on the same sample. It is a genuinely different measurement rather than a better version of the same one.
At Manufacture Versus At Expiry Claims
Here is the distinction that determines what a buyer actually gets.
A count at manufacture, sometimes phrased as at time of production, describes the product as it was blended and encapsulated. A count at expiry, or through end of shelf life, is a guarantee that the stated number will still be present at the end of the printed date when the product has been stored as directed. The second requires a manufacturer to overfill at production by whatever margin their stability data says is needed, and it requires them to have generated that stability data in the first place.
The commercial incentive runs in one direction. An at-manufacture claim allows a larger headline number at a lower cost, because no overfill is required and no long-term stability programme is needed. An at-expiry claim requires putting in more organisms than the label says, testing the product repeatedly across its shelf life, and sizing the shelf life to the data rather than to marketing preference.
| Claim basis | What is guaranteed | Requires overfill | What the consumer receives |
|---|---|---|---|
| At time of manufacture | Count on the day of production | No | An unknown fraction, declining over time |
| At expiry, stored as directed | Count at the end of shelf life | Yes | At least the stated count if stored correctly |
| No timing stated | Nothing defined | Unknown | Indeterminate |
| Refrigerated, at expiry | Count at expiry under refrigeration | Yes | Stated count only if the cold chain held |
The practical reading rule is short. Look on the back for wording about the count being guaranteed through the expiry date or through end of shelf life. If the label says only that a serving provides a given number, without stating when, the claim is undefined, and the honest interpretation is that you do not know what you are buying.
Plate Counting and Its Uncertainty
The precision implied by a probiotic label is far greater than the method can support, and understanding why prevents over-interpreting small differences between products or between test results.
Uncertainty enters at every stage. Weighing and suspending the sample introduces variation, particularly for a heterogeneous powder blend where the organisms may not be evenly distributed through the matrix. Each dilution step carries pipetting error, and because dilutions are sequential, errors multiply rather than average out; a series of several tenfold dilutions accumulates the imprecision of every step before a plate is ever poured.
The plating itself is subject to a fundamental statistical limit. Colonies land on the plate according to a random process, so even a perfect technique produces variation between replicate plates that follows a known distribution. That variation is proportionally larger when few colonies are counted, which is why the convention exists of counting only plates within a defined range, typically between about thirty and three hundred colonies. Below that range, random variation dominates. Above it, colonies merge and are undercounted, and competition for nutrients suppresses growth.
The upshot is that a plate count is properly reported as an estimate with an associated uncertainty, and the uncertainty is not small. Differences between two results on the same product that fall within a factor of about two are generally not meaningful. This has direct consequences for interpreting third-party testing reports: a product testing somewhat below its label claim may be within method uncertainty, whereas one testing at a small fraction of its claim is genuinely failing.
Strain Identity Versus Species Labelling
Bacterial nomenclature runs genus, species, strain, and the properties that matter for a probiotic live almost entirely at the strain level.
A species is a broad grouping, and members of the same species can differ substantially in their genome content. The genes that determine whether an organism adheres to intestinal mucus, tolerates bile, produces particular metabolites, or survives freeze-drying are frequently strain-specific, carried on mobile elements or in variable regions of the genome. Two strains of the same species can therefore behave quite differently in every respect a buyer cares about.
This is why a strain designation matters. It is the alphanumeric code that follows the species name, assigned by whoever deposited the strain in a culture collection, and it is the only identifier that ties a product to whatever work has been done on that organism. A label reading only with a species name is telling you something roughly as specific as naming a breed of dog when the question was about a particular animal’s training.
Identity verification in the laboratory reflects this. Culture-based identification and simple biochemical tests distinguish species imperfectly and cannot distinguish strains at all. Sequencing a conserved gene identifies species reasonably. Establishing strain identity requires whole genome sequencing or a fingerprinting method with enough resolution, and this is what independent testing uses when checking whether a product contains what it claims. Studies of the supplement market have repeatedly found products containing species not on the label, or missing species that were, and mislabelling is more common in products that name only species.
Storage Temperature and Viability Decline
Viability loss in a dried probiotic is a chemical process, and it obeys the same rules as any other chemical degradation: it accelerates with temperature and it requires molecular mobility.
Freeze-drying protects cells by removing water and immobilising what remains in a glassy solid. In that glassy state, molecules cannot move enough to react, and degradation nearly stops. The protective effect depends on staying below the temperature at which the glass softens, and that temperature falls as moisture content rises. This is why the two enemies of shelf life, heat and humidity, are not independent: absorbed moisture lowers the softening point, which means a product that would be stable at room temperature when dry becomes unstable at the same temperature once it has taken up water.
Moisture control is therefore engineering rather than an afterthought. Blister packaging isolates each dose and prevents the repeated humidity exposure that a bottle opened daily experiences. Desiccants in bottles absorb ingress. Aluminium foil barriers outperform plastic films. Capsule shell material matters, because gelatine and cellulose shells contain and exchange different amounts of water with their contents.
Temperature effects are straightforward in direction and large in magnitude. Refrigeration substantially extends viability for most strains, and elevated temperatures shorten it sharply. The implication for real supply chains is uncomfortable: a product may be manufactured, tested and shipped correctly and then sit in a warm warehouse or a delivery vehicle for a period long enough to consume much of its stability margin. A refrigerated product that has been through an interrupted cold chain carries a claim that no longer applies, and nothing about its appearance reveals this.
Survival Through Stomach Acid
Between swallowing and the intestine lies an environment specifically evolved to kill ingested bacteria, and how a product handles it is the subject of a great deal of marketing and rather less clarity.
Fasting stomach contents are strongly acidic, sufficient to kill most vegetative bacteria within a reasonably short exposure. Bile salts released into the small intestine present a second challenge, disrupting bacterial membranes. Any organism reaching the colon in a viable state has survived both.
Several factors modify the challenge, and one of them is under the consumer’s control. Food buffers stomach acid considerably, and the presence of a meal raises stomach pH for a period, so taking a probiotic with or shortly before food generally improves survival compared with taking it on an empty stomach. Fat and protein in the meal enhance this effect. The specifics of ideal timing vary by product, which is why manufacturer directions differ, but the general principle that an empty acidic stomach is the harshest condition holds broadly.
Formulation approaches address the same problem. Enteric coatings resist dissolution at low pH and release their contents at the higher pH of the small intestine, and they work, though they add cost and can complicate dose timing. Microencapsulation embeds cells in a protective matrix. Spore-forming organisms require none of this, since the spore coat is already resistant, and their germination in the intestine is part of their design.
Questions Worth Asking About a Label
A short list of questions separates most well-constructed probiotic products from the rest, and none of them requires specialist knowledge to ask.
Is the count guaranteed at expiry or only at manufacture? This is the first question and the most informative, and the answer is usually on the back panel in small type or absent entirely. Absence is itself an answer.
Are strains identified, not just species? Look for an alphanumeric designation after each species name. A product listing several species with no strain codes has not told you what is in it in any meaningful sense.
Is the count given per strain or only as a total? A total across many species conceals whether any individual strain is present at a level that anything has ever been documented for. Multi-strain products with impressive totals and no breakdown are common, and the breakdown is what carries the information.
How is it packaged and stored? Blister packs and desiccant-protected bottles indicate that moisture was considered. A refrigeration instruction is a real constraint that has to hold through distribution and at home, and a product requiring it that was shipped in summer heat has an unverifiable claim.
Does the number make sense for the intended use? Higher counts are not automatically better. What has been documented for a given strain applies at whatever amount was used, and a much larger dose is a different proposition rather than a stronger version of the same one. A very large headline count across many unnamed strains is a marketing number more often than a functional one.
The underlying discipline is the same one that applies to any supplement label. The prominent number is the one the manufacturer chose to feature, and the informative details are the ones that constrain it: when it was measured, which organisms it refers to, and under what storage conditions it still holds. A product that answers all three plainly is telling you something. One that offers a very large number and answers none of them is telling you something too.
Frequently asked questions
Does a higher CFU count mean a better product?
Not in itself. A count is a quantity of living organisms and says nothing about which organisms they are or whether they do anything useful, and the amount that has been documented for a given strain is a property of that work rather than a threshold that more is always better than. A large total spread across many species, none identified by strain, is generally less informative than a smaller count of a well-characterised strain given at the amount it was studied at. The timing of the count matters more than its size: a modest number guaranteed at expiry may deliver more living cells than a large number measured at manufacture.
Why do independent tests sometimes find fewer organisms than the label claims?
Several reasons contribute, and they are not all misconduct. Plate counting carries real uncertainty, so differences within roughly a factor of two may reflect the method rather than the product. Products labelled on an at-manufacture basis will legitimately test lower once they have aged. Storage during distribution and retail affects viability substantially, and a product that spent time somewhere warm will test below its specification through no fault of its formulation. Genuine shortfalls also occur, and the pattern that indicates one is a result at a small fraction of the claim rather than modestly below it.
Do probiotics need to be refrigerated?
It depends entirely on the strain and the formulation. Spore-forming organisms are stable at ambient temperature by nature, and many freeze-dried products are formulated and packaged specifically to be shelf-stable, with moisture barriers and desiccants doing the work refrigeration would otherwise do. Where a label specifies refrigeration, that instruction is part of the claim and the guaranteed count does not apply without it. Refrigeration will not harm a shelf-stable product and generally extends its viability, so following the label and erring towards cool dry storage is a reasonable default.
What does the strain code after the species name actually mean?
It identifies one specific isolate, usually deposited in a recognised culture collection, and it is the only label element that connects a product to whatever has been documented about that organism. Members of the same species differ genetically in ways that determine adhesion, bile tolerance, metabolite production and survival through drying, so effects established for one strain cannot be assumed for another sharing the species name. The code is often a manufacturer or collection abbreviation followed by numbers, and its presence indicates that the producer knows and can verify which organism they are selling.
Is it better to take a probiotic with food or on an empty stomach?
For most products, with or shortly before a meal is preferable, because food buffers stomach acid and raises the pH for a period, which is the harshest barrier the organisms face. Meals containing fat and protein enhance this buffering. Exceptions exist: enteric-coated products are designed to pass the stomach intact regardless, and spore-based products do not need the protection. Where a manufacturer gives specific timing directions, those directions usually reflect testing on that particular formulation and are worth following over general advice.
This is education, not medical advice. Laboratory results only carry meaning alongside your symptoms, history and examination. Talk to a qualified clinician about your own results before changing anything about your care or supplements.
