The Centenarian Microbiome: What Science Says and What the Industry Sells

No probiotic or postbiotic supplementation has, to date, demonstrated a causal effect on human lifespan. The microbial signatures identified in centenarians (Akkermansia muciniphila, Bifidobacterium longum, Faecalibacterium prausnitzii) reflect statistical associations, and the published clinical benefits of these bacteria apply to specific strains, in specific forms, rarely available in commercial products. This article explains why we chose not to include these three bacteria in the Singular formula, despite their growing visibility in longevity marketing.

What the data on centenarians actually show

Researchers have been sequencing the gut microbiota of centenarians for roughly fifteen years. A 2024 meta-analysis of eight cohorts of long-lived populations (Japan, Sardinia, China, South Korea, Italy, South America) identified microbial signatures shared by individuals who had passed the age of one hundred (PubMed). Two species emerge consistently: Akkermansia muciniphila and Bifidobacterium longum (subspecies longum). Their abundance declines with age in the general population and is maintained, or even increases, in centenarians.

The underlying mechanisms are partly documented. Akkermansia muciniphila carries a thermostable surface protein (Amuc_1100) that binds to TLR2 receptors on intestinal epithelial cells and stabilises tight junctions, the structures that seal the space between mucosal cells (PubMed). Bifidobacterium longum subsp. longum modulates the local immune response via its cell-wall peptidoglycan and influences the gut-brain axis through tryptophan metabolism (PubMed). Bacillus subtilis, a transient environmental spore-former, produces short-chain fatty acids, menaquinone-7 and antimicrobial peptides during its intestinal passage, limiting colonisation by some opportunists.

The most cited clinical trial remains Depommier et al., published in Nature Medicine in 2019. In thirty-two overweight or obese adults, twelve weeks of supplementation with pasteurised Akkermansia muciniphila (strain MucT) improved insulin sensitivity by 28.6% and lowered fasting insulin by 34% compared with placebo (PubMed). A more recent randomised trial in adults over sixty showed that pasteurised Akkermansia (strain HB05P) improved muscle strength and functional performance over twelve weeks (PubMed). For Bacillus subtilis, the reference trial in seniors (Lefevre 2015, strain CU1) documented an increase in secretory IgA levels and a reduction in the duration of winter respiratory infectious episodes (PubMed). For tyndallised Bifidobacterium longum subsp. longum (HT-ES1, strain CECT 7347), a twelve-week trial in two hundred adults significantly reduced the severity of functional digestive symptoms (PubMed).

These data are interesting. They do not prove that supplementing with these bacteria extends lifespan. This is where science and marketing start to diverge.

Three gaps between the studies and the market

Gap 1: the exact strain

Probiotics are strain-specific. Two strains of the same species can show radically different clinical profiles, a principle recognised by the World Gastroenterology Organisation and the International Scientific Association for Probiotics and Prebiotics. Documented benefits therefore apply to the MucT strain (ATCC BAA-835) or its derivative HB05 for Akkermansia, to the CECT 7347 strain for tyndallised Bifidobacterium longum, and to the CU1 strain for Bacillus subtilis in seniors.

On the B2B market, many raw materials labelled "tyndallised Akkermansia muciniphila" or "postbiotic Bifidobacterium longum" do not disclose the strain or use a strain that has not been studied clinically. The finished product can carry the species name while implicitly claiming effects published for a different strain. That extrapolation is not scientifically valid.

This is changing. Some suppliers now disclose their strain identifier and provide a genome-level comparison with the reference strain (Average Nucleotide Identity, or ANI). An ANI of 99.9% with MucT (ATCC BAA-835) implies functional equivalence at the genome level, including for the genes encoding the Amuc_1100 protein. The strain-specificity problem then becomes solvable. But it shifts the critical question onto the manufacturing process.

Gap 2: the exact form and subspecies

Form matters as much as strain. The Depommier 2019 trial used pasteurised Akkermansia (a single thermal cycle at around 70 °C). The live and the pasteurised forms produce comparable effects in that study (PubMed), but this does not automatically validate the tyndallised form (repeated thermal cycles), which has not been studied clinically in the same indications. The Amuc_1100 protein is thermostable up to around 70 °C: the effect of repeated cycles on its activity remains to be documented for each industrial tyndallisation process.

More worryingly, some raw materials sold as "tyndallised" actually use high-pressure steam sterilisation at 115-121 °C for 15 to 30 minutes, autoclave conditions that go far beyond the documented thermostability of Amuc_1100. In those cases, even a strain genetically equivalent to MucT is not enough: the process likely destroys the surface activity on which the clinical effect relies. Scrutiny therefore bears less on the commercial label ("tyndallised", "postbiotic") than on the actual thermal conditions disclosed in the supplier's technical data sheet.

The case of Bifidobacterium longum is even more telling. The species is divided into three subspecies, two of which are commercialised. B. longum subsp. longum is the adult and centenarian subspecies. B. longum subsp. infantis is the dominant subspecies of breastfed infants, specialised in the digestion of human milk oligosaccharides. These are two distinct functional profiles. A product labelled "Bifidobacterium longum subsp. infantis tyndallised" cannot claim the benefits of HT-ES1 studies (which relate to subsp. longum), nor the centenarian signature (also subsp. longum). Yet some B2B raw materials silently perform this taxonomic switch.

Gap 3: association is not causation

The microbial signatures of centenarians are correlations. People who live to one hundred harbour Akkermansia and B. longum because they have lived long in good health, with a varied diet, regular physical activity and managed stress. Their microbiome is probably as much a biomarker of their lifestyle as an independent driver. No randomised controlled trial lasting more than five years has demonstrated that probiotic supplementation extends human lifespan. Mendelian randomisation studies are beginning to suggest a potential causal relationship for Akkermansia, but these data remain preliminary.

Colonisation itself is uncertain. Work published in Cell in 2018 showed that most supplemented probiotics persist in the intestinal mucosa only transiently, with highly personalised host-dependent colonisation resistance (PubMed). The effect therefore depends on continuous intake, without established proof that this translates into a longevity gain.

Our choice for this formula

We reviewed the state of raw materials available in 2026. For each of the three bacteria, we found a gap between what the published studies show and what the B2B market offers. Either the exact strain is not sourceable at industrial scale, or the commercialised form does not match the form studied, or the producer does not disclose the strain identifier. Under these conditions, including these bacteria in a formula positioned on scientific precision amounts to attaching a species name to data that do not apply to it.

We have therefore chosen to keep a single microbiome-related active: GOS (galacto-oligosaccharides). This is not a bacterium, it is a prebiotic fibre. Its value is threefold. It feeds the bacteria each person already has, in particular their resident Bifidobacterium, so the effect does not depend on uncertain colonisation. It is dose-dependent and reproduces its bifidogenic effect across many controlled trials. It is stable, compatible with any diet, and raises no issue of strain, form or regulatory status.

We continue to track the literature. If a pasteurised Akkermansia muciniphila raw material of MucT strain, clinically documented and covered by the Novel Food authorisation, becomes available in B2B at acceptable industrial terms, we will reassess. If the causal evidence on probiotic supplementation and lifespan strengthens beyond cohort associations, we will reassess. Until then, documented restraint seems preferable to claim-led generosity.

Research on the microbiome and longevity is a field under construction. It deserves better than shortcuts that attach the benefits published for a specific strain, in a specific form, under specific conditions, to a species name. Distinguishing what science says from what the industry sells is not self-flagellation. It is the condition for honest formulation.

Frequently asked questions

Why does Singular not include Akkermansia in its formula?#[ + ]

Because the published benefits (Depommier 2019, Kang 2024) relate to the pasteurised MucT strain, covered by the EU Novel Food authorisation 2021/2334. Most B2B raw materials on the market are tyndallised (a different thermal process) and do not disclose the strain. We prefer not to claim clinical effects that do not apply to the strain actually used.

Does the centenarian/microbiota association prove a cause-and-effect relationship?#[ + ]

No. Cohort studies show a statistical association. Centenarians harbour these bacteria because they have lived long in good health, not the other way round. No long-term randomised controlled trial has yet demonstrated that supplementing with Akkermansia, Bifidobacterium or Bacillus extends human lifespan.

Is a good diet enough to support the gut microbiota?#[ + ]

For most people, yes. A randomised trial conducted at Stanford and published in Cell in 2021 showed that a diet rich in unpasteurised fermented foods increases microbial diversity and lowers nineteen inflammatory markers, a result no probiotic supplementation has reproduced to that extent. Artisanal kefir, raw sauerkraut, kimchi, natto, legumes and varied fibres cover the vast majority of needs.

What is the difference between a probiotic and a postbiotic?#[ + ]

A probiotic is a live micro-organism that temporarily colonises the gut. A postbiotic is an inactivated bacterial cell (by heat or pressure) whose surface structures retain their ability to interact with the immune system. Both are valid approaches, but clinical effects are strain-specific and form-specific: data from a pasteurised strain cannot be extrapolated to the same strain tyndallised, let alone to a different strain.

Why do you keep GOS in the formula?#[ + ]

Because GOS (galacto-oligosaccharides) does not rely on uncertain colonisation. It is a fermentable substrate that feeds each person's resident bacteria, stimulates colonic butyrate production, and has a documented dose-dependent bifidogenic effect. The effect is proportional to dose, measurable, and compatible with any diet.

---

References

1. Chen S et al. Consistent signatures in the human gut microbiome of longevous populations. Gut Microbes. 2024;16(1). (PubMed)
2. Depommier C et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nature Medicine. 2019;25(7):1096-1103. (PubMed)
3. Zeng SY et al. Potential Effects of Akkermansia muciniphila in Aging and Aging-Related Diseases: Current Evidence and Perspectives. Aging and Disease. 2023;14(6):2015-2027. (PubMed)
4. Kang CH et al. Pasteurized Akkermansia muciniphila HB05 (HB05P) Improves Muscle Strength and Function: A 12-Week, Randomized, Double-Blind, Placebo-Controlled Clinical Trial. Nutrients. 2024;16(23):4037. (PubMed)
5. Ashrafian F et al. Comparative effects of alive and pasteurized Akkermansia muciniphila on normal diet-fed mice. Scientific Reports. 2021;11:17898. (PubMed)
6. Srivastava S et al. A randomized double-blind, placebo-controlled trial to evaluate the safety and efficacy of live Bifidobacterium longum CECT 7347 (ES1) and heat-treated Bifidobacterium longum CECT 7347 (HT-ES1). Gut Microbes. 2024;16(1). (PubMed)
7. Ku S et al. The role of Bifidobacterium in longevity and the future of probiotics. Food Science and Biotechnology. 2024;33(9):2097-2110. (PubMed)
8. Lefevre M et al. Probiotic strain Bacillus subtilis CU1 stimulates immune system of elderly during common infectious disease period. Immunity & Ageing. 2015;12:24. (PubMed)
9. Zmora N et al. Personalized Gut Mucosal Colonization Resistance to Empiric Probiotics Is Associated with Unique Host and Microbiome Features. Cell. 2018;174(6):1388-1405.e21. (PubMed)
10. Wastyk HC et al. Gut-microbiota-targeted diets modulate human immune status. Cell. 2021;184(16):4137-4153.e14. (PubMed)