Self & Other · The Microbiome Decade

Babies, Briefly

The microbiome and autism. Some studies have found differences in the gut microbiomes of autistic and non-autistic children. Some have found correlations between certain microbiome features and the severity of certain co-occurring symptoms (gastrointestinal issues, in particular).

What is dangerous is the leap from these findings to claims that autism can be treated, prevented, or cured through microbiome interventions. There is no good evidence for any such intervention. Some of the protocols promoted on the basis of these claims involve aggressive treatments, including unauthorized fecal transplants, that carry real risks for children. Autism is a complex developmental difference with strong genetic components and a long history of being targeted by unscientific cures. The microbiome research is interesting and worth following. It is not, in 2025, a treatment.

The microbiome and longevity. Studies of centenarians have found certain microbiome features that distinguish them, on average, from younger old people. Some of these features may reflect features of long-lived people that affect their microbiomes, rather than features of microbiomes that produce long lives.

What is overhyped is essentially the entire industry that has grown up around selling longevity-via-gut-bacteria products. None of these products has been shown, in serious trials, to extend life. The actual interventions that have been shown to extend life involve the things we have been discussing throughout this book — sleep, movement, social connection, not smoking, getting your medications and your vaccinations — and microbiome supplements have not, so far, been added to that list.

Babies, Briefly

One area of microbiome research that has produced both real findings and substantial public confusion is early life.

Babies are born essentially without their adult microbiome. The first weeks and months of life are when the gut is colonized for the first time, in a process that involves bacteria from the birth canal (in vaginally delivered babies), the skin and breast (in breastfed babies), and the surrounding environment. The pattern of colonization in the first year of life sets up a community that, while it changes throughout life, has lasting features.

Several real associations have been found. C-section delivery is associated with somewhat different early microbiomes than vaginal delivery, and some studies have found higher rates of certain conditions (asthma, certain allergies) in children delivered by C-section, though the size of the effect is modest and other factors (including the antibiotics often given during C-sections) confound the picture. Breastfeeding produces a microbiome substantially different from that of formula-fed babies, with breastfed infants having higher proportions of certain bacteria that appear to be beneficial.

What has been overinterpreted is the implication that mothers who deliver by C-section or who use formula are damaging their children’s microbiomes in ways that have major lifelong consequences. Some entrepreneurs have promoted “vaginal seeding” — swabbing C-section babies with maternal vaginal fluid — as a way to address the supposed damage. The practice has not been shown to provide benefit, has been associated with rare but real risks of transferring pathogens like Group B strep, and is not recommended by major medical bodies.

The reasonable summary is that mode of delivery and feeding do affect the early microbiome, and these factors have small effects on certain later outcomes. The effects are real and worth understanding. They are also small enough that they should not be a source of guilt for any parent whose circumstances led to a different choice. Babies are, on the whole, remarkably robust to variation in early microbiome composition, and the major factors that affect a child’s life are, mostly, not microbiome-related.

What About Microbiome Tests?

A whole industry has emerged offering at-home microbiome testing kits. You collect a stool sample, send it in, and receive a report listing the bacteria found in your gut, often with recommendations for diet, supplements, and lifestyle changes designed to “optimize” your microbiome.

The technology behind these tests is real — the sequencing is done with the same kinds of methods used in research labs — but the interpretation is mostly not. There is, currently, no consensus on what an optimal microbiome looks like for any given person. The recommendations these companies provide are based on associations that are often weak or contested, and the recommendations themselves often boil down to “eat more fiber, eat more fermented foods, sleep well, manage stress” — advice you do not need a stool sample to receive. Some tests recommend specific supplements that the same company sells.

If you are curious about your microbiome composition, these tests can satisfy that curiosity. As tools for changing your health in ways the recommendations support, they are mostly oversold. Save the money for something with better evidence behind it.

What the Next Decade May Bring

The first decade of intense microbiome research produced more associations than interventions and more hype than therapies. The next decade is likely to produce more therapies, fewer associations, and — ideally — less hype. Several specific directions are worth watching.

Engineered probiotics. Rather than the current generation of products, which mostly contain bacteria selected for survival in shipping, the next generation of probiotic products may contain bacteria specifically engineered to perform specific functions in the gut — producing particular metabolites, displacing particular pathogens, modulating particular immune cells. Several such products are in development. They will probably be regulated as drugs rather than as supplements, which means they will have to demonstrate efficacy in clinical trials before being approved. This is the kind of microbiome-based therapy that may actually deliver on the promises the supplement industry has been making for a decade.

Standardized fecal transplant products. The C. diff use case has produced enough demand that several companies are now producing carefully screened, freeze-dried, capsule-form fecal transplant products approved as medical treatments. These are likely to be expanded to other indications as the evidence develops.

Microbiome-tuned cancer immunotherapy. As the link between gut bacteria and immunotherapy response is worked out, oncologists may begin to deliberately modify patients’ microbiomes as part of cancer treatment. Whether this works in practice, and at what scale, remains to be seen. The early results are promising enough that the trials are real.

Better measurement. Most microbiome research so far has identified bacterial communities by sequencing fragments of bacterial DNA. The next generation of techniques is starting to look at what those bacteria are actually doing — which genes they are expressing, which metabolites they are producing, how they are interacting with each other and with the host. This is harder, more expensive, and more informative. The papers that come out of this work over the next decade will probably be more useful than the papers we have had so far.

The honest summary of the microbiome decade is that the science is real and remains exciting, that several genuine clinical applications have emerged, that the hype substantially exceeded the evidence, and that the next decade of work will probably narrow the gap between what is claimed and what is true.

Your microbiome matters. The community of bacteria living in your gut is not a passenger; it is a working part of you that contributes to your immune system, your metabolism, and probably several other things we have not yet fully mapped. Eating fiber feeds it. Antibiotics damage it. Stress, sleep, and other systemic factors shift it. Caring about your microbiome, in the form of generally caring about a sensible diet, is reasonable.

What is not reasonable, on the available evidence, is the elaborate edifice of products, protocols, and personalized recommendations that has been built on top of the underlying science. Most of it does not deliver. Some of it does. Distinguishing one from the other is a skill that requires looking at evidence rather than at marketing, and the same skill that has served you throughout this book — noticing what is well-supported, what is suggestive, and what is being sold to you on premises that the science does not support — will continue to serve you here.

The next chapter of this book is the glossary. It is a short reference for the terms that have appeared in the previous chapters and that will appear in any medical news article you read for the rest of your life. After that, the epilogue, where we return to the seven-year-old on the bicycle.

Almost there.

All 21 chapters of Self & Other · Library · All courses