Zoh Science · Milk kefir
A jar of kefir is a tiny community at work. Bacteria and yeasts help transform milk—but their names, their numbers and their effects on your health are three different stories. Here is how to tell them apart.
Explain it simply
The tiny version
Imagine a kitchen where different cooks do different jobs. Some change the ingredients. Others use what the first cooks leave behind. Together, they make something no one cook could make alone.
Kefir microbes can work a little like that. Researchers have observed microbes sharing resources during fermentation. That helps explain how a community can thrive in milk. [1]
The analogy stops here: a busy kitchen does not tell you whether its food will treat a health condition. We need studies in people for that.
A changing community inside the jar
Grain-fermented kefir communities can include lactic acid bacteria, acetic acid bacteria and yeasts. The organisms and their proportions vary; these groups are a way to understand the community, not a certificate for every jar or sachet. [1] [2]
Different jobs, one fermentation
It helps to begin with what the microbes do. Lactic acid bacteria contribute to acidification; other bacteria and yeasts participate in a wider network of fermentation activity. These categories give us a map of the work taking place. They do not tell us the exact membership of an individual jar.
In a study of three kefirs, researchers followed both microbial changes and flavour compounds over 24 hours. Lactobacillus kefiranofaciens was prominent early, while Leuconostoc mesenteroides became more prominent later. Acetobacter pasteurianus abundance was associated with acetic acid. These observations connect particular communities with the chemistry of a fermentation; they are not a recipe for selecting a health benefit. [2]
Think of the jar as a process unfolding through time, rather than a photograph of a fixed team. The question “what is in kefir?” becomes more useful when we add: which kefir, sampled where, at what stage, and measured how?
Why the grain and the drink can tell different stories
Blasche and colleagues examined how a kefir community organised itself between the grain and the surrounding milk. Microbes occupied these environments differently, and organisms entering the milk changed the resources available to later arrivals. The study helps explain why looking at isolated organisms does not capture the whole fermentation. [1]
This distinction matters when reading a headline about “kefir microbes”. A finding about a grain is not automatically a measurement of the drink poured from it. Before treating a species list as an ingredient list, check what the researchers actually sampled.
Why there is no universal microbe count
A 2023 study analysed kefirs prepared from 64 grains originating in 25 countries. It found several community types, with different patterns of dominant bacteria. This is a useful demonstration of variation under the study's preparation conditions; it does not establish that every commercial starter or home batch has the same composition. [6]
How researchers look also matters. Marsh and colleagues used sequencing to examine bacteria and fungi in 25 kefir milks and their associated grains. Sequencing and cultivation ask related but different questions: detecting microbial DNA is not the same measurement as growing viable organisms under chosen laboratory conditions. [7]
Imagine one list describing a grain, another describing a finished drink, and a third combining findings from several countries. Adding those lists together may produce a large number, but it does not show that all those organisms share your glass. Nor does counting species tell us how many distinct strains have been identified.
Four labels that should never be mixed up
A named group, such as Leuconostoc mesenteroides.
Does not establish: Which particular strain was tested.
A more specific identity, often with an extra code.
Does not establish: That every strain of the species has the same clinical effect.
A viability measurement under a stated method; CFU means colony-forming units.
Does not establish: How many different species or strains are present.
What a defined intervention did in a particular study.
Does not establish: That any product with a similar microbial name will do the same.
A useful probiotic claim needs more than an impressive list: identity, an adequate amount and evidence of benefit matter. [3]
Explain it like I’m a clinician
The clinical lens
Separate ecological function from clinical efficacy. Blasche and colleagues mapped community dynamics and metabolite exchange in grain-fermented kefir. Organisms were distributed differently between grain and milk, and nutrient sharing helped sustain the community. This is mechanistic fermentation evidence—not a trial of patient outcomes. [1]
Walsh and colleagues combined sequencing, volatile-compound analysis and experimental manipulation across three kefirs. This supports studying community–flavour relationships; it does not validate a therapeutic dose or a universal strain inventory. [2]
A small correction with a big lesson
A 2017 correction clarified that a Leuconostoc mesenteroides isolate used in the Walsh study should be designated DPC 7047, rather than 213M0. A species name alone would hide that difference. We link the correction alongside the paper. [2a]
For a clinical claim, ask: Which preparation? Which population? What amount and duration? Compared with what? Which measured outcome? A microbial mechanism is a reason to investigate, not the answer to those questions.
What it means for my health
What it means for your health
A small lactose-digestion trial.
An acute response, not a long-term result. Study [4]
In a 2003 study, 15 adults with lactose maldigestion received test meals containing 20 g lactose from different dairy preparations. Plain kefir produced less breath hydrogen than milk; kefir preparations also reduced perceived flatulence severity relative to milk. Responses were measured over eight hours. [4]
What that means: some people may handle a studied kefir preparation differently from milk. What it does not mean: all kefir is lactose-free, milk allergy has been addressed, or the results prove a long-term “gut reset.” It was not a trial of Zoh starter.
A 2023 systematic review covering 16 studies found that 12 had an overall high risk of bias. Its search ended in July 2021, so it is a useful appraisal of that evidence base—not a complete account of all later research. [5]
No health score for a microbe count: the papers discussed here do not establish that a kefir with more named species is better for your health. A larger number is not a treatment recommendation.
How to use this information
When choosing a culture
Ask what a microbial list represents. Is it information about the actual product, a tested finished batch, or organisms described somewhere in kefir research? Those are different kinds of evidence. If a seller gives a count, look for the unit and what was measured. A number without that context cannot answer a health question.
When making a batch
Follow your culture's instructions and keep a simple record of the milk, volume, temperature and fermentation time. If the result changes, those notes make troubleshooting more useful. Extra sourness or a thicker texture cannot tell you which strain is present, and extending fermentation is not a reliable way to engineer a particular health effect.
When reading a promising headline
Translate it into a question you can check: did researchers study a microbe in a laboratory, a fermented food, or people consuming a defined preparation? Then look for the outcome. A change in a laboratory measurement and an improvement in a person's symptoms are not interchangeable. This lets you stay curious without asking one study to answer more than it tested.
If you enjoy kefir, think of it first as a fermented food. For a specific symptom or diagnosed condition, discuss the relevant evidence with your clinician rather than choosing a treatment from a list of microbial names.
What this says about Zoh
These studies examine particular research kefirs or preparations. They do not certify the microbial composition, live count or clinical effects of Zoh's sachets or your home batch. Product-specific claims need product-specific documentation.
Quick questions
Does every kefir contain 59 strains?
No universal count follows from these studies. Adding together organisms reported across different samples does not show that they all occur in one drink. A species list is also not a strain count.
Can I identify microbes from a photo?
No. The cover is a conceptual illustration, not microscopy or a strain inventory. A kitchen photo cannot establish microbial identity or a viable count.
Are all microbes in a fermented food automatically probiotics?
No. The probiotic concept includes a demonstrated health benefit at an adequate amount. Simply being present in a fermented food does not establish that evidence for every organism. [3]
Keep exploring
- Lactic acid bacteria: what they do in kefir
- How to read kefir health claims
- Practical help with your batch
- Zoh milk kefir starter and current instructions
Research notes: follow the claim
Selected references for this explainer, not a systematic review. Fermentation experiments, human trials and expert consensus are labelled separately. No independent clinical review is claimed.
- Fermentation experiment. Blasche S et al. Metabolic cooperation and spatiotemporal niche partitioning in a kefir microbial community. Nature Microbiology. 2021;6:196–208. DOI and paper. Supports community dynamics and metabolite exchange; not a health-outcome trial.
- Fermentation experiment. Walsh AM et al. Microbial Succession and Flavor Production in the Fermented Dairy Beverage Kefir. mSystems. 2016;1:e00052-16. DOI and paper. Supports the specific fermentation and flavour observations above. [2a] Correction: 2017;2:e00003-17. Corrected strain designation.
- Expert consensus. Hill C et al. ISAPP consensus statement on the scope and appropriate use of the term probiotic. Nature Reviews Gastroenterology & Hepatology. 2014;11:506–514. DOI and statement. Defines the evidence context for probiotic terminology.
- Human trial. Hertzler SR, Clancy SM. Kefir improves lactose digestion and tolerance in adults with lactose maldigestion. Journal of the American Dietetic Association. 2003;103:582–587. DOI · Study abstract. Small acute study; does not establish effects for every kefir preparation.
- Systematic review. Kairey L et al. The effects of kefir consumption on human health: a systematic review of randomized controlled trials. Nutrition Reviews. 2023;81:267–286. DOI · Abstract and limitations. Search through July 2021.
- Community profiling. Walsh LH et al. Analysis of the milk kefir pan-metagenome reveals four community types, core species, and associated metabolic pathways. iScience. 2023;26:108004. DOI and paper. Study of kefirs made from 64 grains; not a clinical trial.
- Sequencing study. Marsh AJ et al. Sequencing-based analysis of the bacterial and fungal composition of kefir grains and milks from multiple sources. PLoS ONE. 2013;8:e69371. DOI and paper. Sample and measurement context for interpreting microbial lists.
Keep the story going
