Not all probiotics are the same: species, function, and dose matter

The word "probiotic" gets stretched across an enormous range of very different things. A cup of yogurt, a jar of kimchi, a bottle of kombucha, and a capsule of eight carefully selected strains are all technically sources of live bacteria that can be called probiotic. That is a real category, and every one of those foods and formulas has a place.

It is also a category as broad as the word "vehicle." A bicycle and a fire truck are both vehicles. They do very different jobs. Whether the one in front of you is the right tool for the moment depends on details the category name does not tell you.

For probiotics, the details come down to three questions. All three matter, and asking any one of them alone is not enough:

1. Species. Is the underlying biology the kind that could plausibly do this job at all?
2. Strain. Which member of that species is inside, and what has *it* specifically been shown to do?
3. Dose. Is enough of it arriving in the gut to actually be an intervention, rather than a rounding error against the hundred trillion bacterial cells already living there?

Fermented foods clear the first question easily for general gut support; a job they have been doing across human cultures for thousands of years, and doing well. A targeted probiotic supplement is aiming at a narrower question, one the second and third filters are built for. The two are complementary, not competing.

Here is how each of those three filters actually works.

Species vs strain vs blend

Three words get used almost interchangeably in probiotic marketing, and it is worth slowing down on the differences.

A species is the broad biological group. Lactobacillus plantarum is a species. So is Bifidobacterium longum. When a label lists a species and stops there, it is telling you the neighborhood, not the address.

A strain is a specific, genetically distinct member of a species. Lactobacillus plantarum DT88 is a strain. Lactobacillus plantarum DT22 is also a strain. Both are the same species. In the lab, they behave like two different organisms.

A blend is a curated mixture of specific strains chosen to do something together that any one of them cannot do alone.

Here is the analogy that makes this stick. Species is like the breed of a dog. Strain is the individual dog. Every golden retriever is a golden retriever. That does not mean every golden retriever will fetch, or swim, or get along with your cat. Two dogs of the same breed can have very different temperaments, jobs, and skills.

The same is true, in a much more literal way, for bacteria.

Surface chemistry: why one strain binds and another does not

Bacteria interact with the world through their outer surface, such as proteins, sugars, and a sticky matrix the cell secretes around itself. The exact recipe of that surface varies from strain to strain, even inside the same species.

That matters because "surface chemistry" is not an abstract idea. It determines whether a bacterium sticks to the gut lining or slips past it. Whether it can survive a rough trip through the stomach. Whether it can crowd out an unwelcome microbe. And, in newer work, whether it can hold onto a passing microplastic particle inside the gut.

In one 2025 study, researchers screened hundreds of probiotic strains for their ability to bind polystyrene plastic particles. The results were striking. Two strains of the same species differed by more than thirteenfold in how much plastic they held onto. One strain, DT88, bound roughly 80% of the particles offered to it. Another strain from the same species bound about 6%.[1] Same species. Same category on a label. Very different behavior.

That is the second filter, in a single dataset. Species matches. Function does not.

How much actually reaches the gut

The third filter is dose, and it is the one most easily overlooked because it does not show up on a label as a headline number.

The adult gut already carries something on the order of a hundred trillion bacterial cells. Any incoming probiotic has to arrive in high enough numbers to be more than a rounding error against that background, and it has to arrive in the strains that actually do the job you care about.

Fermented foods contain live cultures at counts that vary by orders of magnitude between batches, temperatures, and how long the jar has been open. That variability is part of their character, and it is not a problem for what those foods are traditionally asked to do. It is worth knowing about when the question shifts from general gut support to a specific mechanism.

The species mix is also worth knowing. Yogurt cultures are chosen for the job of turning milk into yogurt reliably. Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus are excellent at that. Kimchi is wilder and denser, built around Leuconostoc, Weissella, Lactobacillus sakei, and others, in ratios that shift as the jar ages. Both are wonderful in what they contribute. Neither was assembled with polystyrene binding in mind, and neither delivers a predictable dose of a specific strain identified for a specific job.

That predictability (a defined strain, at a defined dose, arriving in the gut in a defined form) is what a targeted probiotic supplement is built to provide. It is a different job than fermentation, and it lives comfortably alongside the fermented foods you already eat.

Why curated multi-strain formulas exist

If strain matters this much, one obvious question follows. Why not just pick the single best strain and call it done?

Because the gut is a community, and communities do things individuals cannot. Different strains bring different surface features, different metabolic products, and different tolerances. A well-chosen blend can cover a wider range of conditions across the length of the gut than any one strain can on its own. In laboratory testing of plastic binding specifically, multi-strain mixtures have sometimes bound more plastic than the arithmetic average of their components would predict, which hints at some form of cooperation between strains rather than simple addition.

That is the case for a formula built as a formula, rather than a single hero organism packaged alone.

Where Winnow fits in

Winnow is a curated eight-strain consortium of lactic acid bacteria and Bifidobacterium. The strains were chosen because they clear all three filters at once: species with the underlying biology to bind, specific strains shown in laboratory testing to actually do so, at a dose designed to reach the gut in the numbers the preclinical work suggests actually matter, alongside the ordinary daily-probiotic considerations you would want from anything you take regularly.

Keep eating the yogurt and the kimchi. They are doing what fermented foods have always done for the gut, and they are doing it well. A targeted probiotic is a different question (a specific strain, at a specific dose, chosen for a specific job) and it sits comfortably next to the fermented foods rather than in place of them.

The next time you see the word "probiotic," treat it like the word "vehicle." Useful category. Not enough information on its own. The interesting questions are always which one, what it was built to do, and how much of it is getting where it needs to go.

For deeper reading on the underlying binding evidence, see our long-form review of the peer-reviewed literature on probiotics and microplastic binding, and our companion piece on whether probiotics survive stomach acid. You may also find our overview of how a gut-binding product could work in the first place useful.

Sources

[1] Teng, X., Zhang, T. & Rao, C. Novel probiotics adsorbing and excreting microplastics in vivo show potential gut health benefits. Front. Microbiol. 15, 1522794 (2025). https://pubmed.ncbi.nlm.nih.gov/39867494/