Everything I know about the Gut Microbiome

Over the past few years I've been reading a lot about our microbiome. What it does, how it works, how (and if) you can change it, why you might want to, and more. Now, I've written it down.

In 2023 I had my appendix removed, which was my first ever surgery and came with massive doses of antibiotics. It was only the second time in my life I’ve had antibiotics. A couple week later, I had a period of a few months where I struggled with really strange symptoms, both gut symptoms, but also heart rhythm disturbances. It was scary, and I noticed that my heart symptoms seemed correlated with gut symptoms. A cardiologist found nothing wrong with my heart, thankfully, so I started pulling on the gut-heart connection. There is nascent research on this, but no real therapeutic options yet. Via a great company in Singapore, called Genixpro, I had my gut microbiome tested, and it turned out to actually be severely depleted at the time.

As I focused on restoring my microbiome from the nuclear antibiotic apocalypse that had happened to it, my heart symptoms gradually resolved. Ever since, I’ve gained an interest in the microbiome and learned a lot about it.

The original premise of this newsletter is that as an insatiably curious and analytical person, I learn lots of things about lots of things, and bring that to you in summarized and synthesized form. So, here is what I’ve learned so far about the Microbiome.

Foundations

Bacteria are single-cell organisms and they are everywhere. On your skin for example, there are millions of bacteria per square centimeter across thousands of species, depending on the location. There are many ways to categorize them. For example, we can look at their shape (Salmonella even looks nasty).

Source: https://biologylearner.com/bacteria-definition-characteristics-examples/

But the most useful way to look at them is by focusing on what they do. Do they produce certain products that affect the body? Do they interact with the body in some direct way? This functional categorization is what we really want to know. All bacteria need energy, and to get it, they break down (metabolize) stuff from their environment and excrete waste products. An important distinction to make is whether a bacterium is aerobic or anaerobic, i.e. does it use oxygen or not for its metabolic process. Then, within aerobic (oxygen using) bacteria some prefer to use oxygen but can do without, and others require oxygen. Likewise, some anaerobic bacteria like Lactobacillus can survive in an environment that has oxygen, but some will immediately die. Now, if we want to understand in detail how bacteria affect the body, there is a ton of cell biology that we need to understand. It’s complicated, and I don’t think it’s super important from a practical standpoint exactly how and why it works. But the key mechanisms at work are these:

1. The bacteria can metabolize things for us, like fiber, and the breakdown products, like Butyrate, can cause downstream effects.

2. The bacteria can have surface elements that bind to receptors in our body, which can cause effects directly.

3. Bacteria can trigger the immune system via various paths; excretion product, surface protein, proteins in the tail, and more.

The gastrointestinal system

Most people will know the basic makeup of the GI system: food starts to get digested already in the mouth, then in the acid environment of the stomach it is broken down further, and then it passes through the small and then large intestine to extract as many nutrients as possible. The total length of the human GI Tract from mouth to anus is about 9 meters! What most people probably won’t know, is that the inside of the gut is not considered to be inside your body by biologists. There is a 9 meter long tube running right through your body, which is from a biological standpoint is part of the outside world. Until food is broken down and actually passes through the gut wall, it’s just flowing through a tunnel where the body can interact with it through a protective barrier. This is critical, because the real inside of your body (blood, cells, plasma, tissue, etcetera) is sterile. And in fact, if bacteria were to actually enter your body, that is not good. That makes the gut lining as critical as our skin to keep the outside world outside, while selectively allowing nutrients in.

Every time I think about this barrier function (both skin and gut lining), it blows my mind that it works almost all of the time. How is it possible that the vast majority of people don’t have violent gut issues? How can your skin aborb moisture and also sweat, but water doesn’t just uncontrollably leak out of your skin? It’s ridiculous.

One critical element to make all this work is the microbiome. Until recently, we barely knew anything about it. With regard to the gut biome, the key reason is that most of the microbiome action happens in the distal (last part of the) colon, where there is a total lack of oxygen. In fact, bacterial density increases dramatically as food moves through the GI tract: the stomach has relatively few bacteria due to its acidic environment, the small intestine has moderate numbers, but the colon is absolutely packed with them - roughly 100 billion bacteria per gram of content, with bacteria ultimately making up about half the weight of stool.

So that obviously means that aerobic bacteria cannot survive there. But more troubling for science has been that the bacteria that do live there will instantly die if they are exposed to oxygen. Therefore, we cannot grow them outside the body with traditional methods, and hence we could not study them.

Biome science is very new

Probiotics like various Lactobacillus strains have been on the market since the 1970s. But is that because they are important? Or because they are available? A key eye-opener is that the only reason they are in so many supplements is because they are easy to grow. Lactobacillus lives in the upper parts of the GI tract, where there is still oxygen. We can easily produce these strains and put them into pills with a stable shelf-life.

But most of the gut bacteria live in the oxygen-free environment much deeper into the GI tract. That’s where fiber is digested and, somehow, turned into ‘actions’ that are extremely beneficial for the human host. It wasn’t possible to learn much more until DNA sequencing became available. That only happened in the 2000s. DNA sequencing let scientists identify bacteria directly from stool samples by their genetic signatures. In 2008 or so, 16S sequencing enabled scientists to identify which bacteria were present, and in which proportions. But it still didn’t tell us what they did. Only around 2015 when something called shotgun metagenomic sequencing became affordable, was it possible to sequence all the genes present in the entire microbial community. Now we finally got to know what the bacteria actually do: Do they produce Butyrate? Break down specific fibers? Produce vitamins? Do they interact directly with the gut wall? As I wrote early-on, this functional knowledge is much more important than knowing which latin names are in there.

So, it’s only been ten years since scientists gained the technology to even understand what gut bacteria do and how they do it, and from there try to figure out what effects that might have on the body, and why that is the case. Progress! But doing science takes time, and translating science into approved, safe and effective therapies takes even more time. As a result, we are really at the dawn of a new era, where our ability to improve lives by modifying microbiomes will take off.

So what have we learned so far

There is lots of things we learned. It’s complicated and a bit magical. I think if you read this, you’ll learn a bunch of interesting biology. I’ll keep it simple.

1. Without our gut bacteria, we could not eat plants.

Humans only have about 20 enzymes capable of breaking down complex carbohydrates. Our gut bacteria collectively have over 30,000 different enzymes to break down fiber. Maybe you didn’t think of fiber as a carbohydrate, but it is. Cellulose is a carb, a ‘polysaccharide’ to be precise, just like starch. The ‘ose’ gives it away: glucose, fructose, lactose, amylose (pasta starch) are all carbs. Our bodies can digest some complex carbs, like amylose, but some are indigestible, like cellulose. The bacteria in our microbiome feed on these fibers and break them down into things our body can handle.

Different bacterial species have different toolkits. Bacteroides species deploy specialized gene clusters called polysaccharide utilization loci, think of them as molecular scissors, each one custom-designed for a specific type of fiber. Bacteroides thetaiotaomicron dedicates 18% of its entire genome to 88 different sets of these scissors. Firmicutes bacteria like Ruminococcus build massive enzyme complexes called cellulosomes that work like molecular cranes to tear apart tough fiber structures.

Then, once the mechanical breakdown to smaller carbohydrates is done, the bacteria start fermenting these carbs.

---

Quick cell biology lesson

If you really hated biology and chemistry in school, you could skip this, but if you bear with me for three minutes you’ll learn some basics of metabolism that will be helpful to frame everything else.

The unit of energy of biology is a molecule called ATP (Adenosine Tri-Phosphate). It can release one Phosphate to become ADP (di-phosphate), and in doing so it releases energy. This is the only molecule the body can use to do any kind of ‘work’. Move a muscle: tiny two-legged machines attached to one muscle fiber consume ATP and ‘walk’ up the adjacent muscle fiber. The body can make ATP in various ways. The most straightforward way is to take Glucose and oxidize (burn) it with Oxygen, using the freed-up energy to re-attach one P to ADP and create ATP + CO2 and water. Let me give you some perspective on the scale. The average human at rest turns over about 1.5KG of ATP per hour, with a cell churning through as many as 10 million ATP molecules per second. A person working out can turn over an amount equivalent to their entire body supply of ATP every two minutes. When I say ‘turn over’, I mean using the energy stored in an ATP to do some work, thereby forming ADP and a free phosphorus group, and then using the energy from glucose to reattach the P to reform ATP¹. This oxidative ‘recharging’ process can also ‘run on’ ketone bodies instead of glucose, which is when someone is ‘in ketosis’ (simplifying a bit here). To complete this section, two last facts to know: one glucose molecule can be burned to recharge 32 ADPs back to ATP, and one single grain of sugar contains 4x10^18 (a 4 with 18 zeroes) glucose molecules.

If our cells temporarily don’t have enough oxygen, they can thankfully still keep working. This happens for example when you’re sprinting. In that case we can also use the energy in glucose to create ATP without using oxygen in a process that releases lactate, which becomes lactic acid. This anaerobic process is called fermentation! I bet you didn’t know that some fermentation happens inside our cells.

Both oxididation and fermentation leads to an excess of H (Hydrogen) atoms. These are highly reactive and need to be ‘dumped’ somewhere in a stable form. If there’s oxygen, we can make nice and simple CO2 and H2O end products. But in fermentation, we don’t have enough O molecules to do that, so we need to find some other hydrogen acceptors. Nature has a few options where we can dump hydrogen:

1. Aldehydes, which can take an H to become alcohols (🍻🍻)

2. Ketones, which become acids

3. CO2, which with enough Hs can form H2O and CH4 (aka Methane gas 💨💨)

To unlock energy from carbohydrates, they can be oxidized (burned) with oxygen which always creates CO2 and water, or they can be fermented without oxygen, which creates a wide range of bigger ‘waste products’ like alcohol or various acids.

Short-chain fatty acids

We are ready now to go back to the guts. When fiber is fermented, ‘short-chain fatty acids’ (SCFAs) are created via the ketone group → acid pathway described above: Acetate, Propionate and Butyrate. These SCFAs provide roughly 10% of our daily calories. But the energy is almost beside the point. These molecules are signaling compounds that affect everything from your colon cell function to your immune system to your brain. We have learned that these SCFAs are the key to most of the weird and magical results that happen when you start messing with microbiomes.

Butyrate deserves special attention. The cells lining your colon (colonocytes) get 60-70% of their energy from butyrate, not glucose like most other cells in your body. When germfree mice lacking gut bacteria are studied, their colonocytes show a 56% reduction in ATP levels and massively reduced mitochondrial function—but only in the colon, not in other organs. The colonocytes literally depend on bacterial butyrate to function properly.

2. The microbiome is a chemicals factory (Vitamins and GLP-1)

Beyond breaking down fiber, we’ve learned that the microbiome produces essential compounds our own bodies can’t make.

Vitamins
Humans have evolutionarily lost the ability to synthesize complex vitamins, outsourcing this to bacteria. A massive analysis of 8,000 gut genomes found that 40-65% of bacterial species can synthesize B vitamins and Vitamin K.

Take Vitamin B12 (cobalamin). It is structurally incredibly complex and human genomes don’t have the machinery to build it. But Bacteroides species in your gut still carry the 20+ genes required to synthesize it from scratch. In fact, Bacteroides species can produce all nine distinct B-vitamins.

The data suggests the microbiome provides anywhere from 27% to 86% of our daily requirements for vitamins like folate, riboflavin, and B12. It’s a symbiotic trade: we feed them fiber, and they synthesize the vitamins we can’t make ourselves.

Metabolic signaling and GLP-1
You’ve probably heard of GLP-1 (the hormone mimicked by Ozempic). It turns out the microbiome plays a role in stimulating your body’s own production of it.

The mechanism is fascinating and involves Bile Acids. The liver makes “primary” bile acids to help digest fats. But when these reach the colon, bacteria (specifically species like Clostridium scindens) transform them into “secondary” bile acids through chemical reactions like deconjugation and dehydroxylation.

These bacterial-modified bile acids bind to specific receptors in your gut lining, such as the TGR5 receptor. When these receptors are activated by the secondary bile acids, they signal your gut cells to secrete GLP-1, which improves insulin sensitivity and regulates glucose homeostasis. Crazy!

3. There is massive microbiome variability between people, but functionally there is huge overlap

People only share about 43% of bacterial species. Your gut bacteria and mine might look quite different at the species level—different Latin names, different proportions. But when researchers looked at what the bacteria can actually do, people shared 82% of functions. Different bacterial species carry similar enzymes. Multiple species can produce butyrate, break down specific fibers, or synthesize vitamins.

Therefore focus has recently turned to focusing on what the bacteria actually do. What feeds them? and what do they turn it into? What matters isn’t so much which species you have, but whether your microbiome collectively has the functional capacity to perform essential tasks: breaking down fiber, producing SCFAs, synthesizing vitamins, transforming bile acids. That changes the game quite a bit, and should make you wonder the next time you see an ad for a supplement with some bifidobacteria.

4. The microbiome can change rapidly

Your gut microbiome responds dynamically to diet, often within days. If you go on vacation to another continent, your microbiome will be very, very different when you come back.

Some examples: when people eat resistant starch, Ruminococcus bromii, the keystone degrader of resistant starch, explodes from 3-4% of the gut population to 17% in just a few days. This single species unlocks the starch, releasing breakdown products that feed dozens of downstream species, which then produce the butyrate that benefits you. Stop eating resistant starch, and R. bromii populations decline again. Different fibers recruit different teams within days. Inulin increases Bifidobacterium adolescentis four-fold. Pectin specifically enriches Eubacterium eligens and Faecalibacterium prausnitzii. Beta-glucans from oats favor Bacteroides uniformis.

Bacteria also mutate, and can pick-up entirely new functions through horizontal gene transfer from other species. Japanese people have Bacteroides strains that can degrade seaweed polysaccharides. These enzymes are completely absent in North American populations. This plasticity is both encouraging and concerning. Encouraging because it means diet changes can rapidly produce changes. Concerning because multi-generational fiber deprivation causes progressive, potentially irreversible loss of fiber-degrading bacteria. Westernization has been associated with depletion of species like cellulolytic Ruminococcus that are common in hunter-gatherers but rare in industrialized populations. We may be losing metabolic capacities we evolved to depend on.

5. Microbiomes also metabolize some random things, like certain drugs

Here’s where things get weird. Your gut bacteria don’t just break down food, they also metabolize drugs in ways that can completely change how medications work.

The first example we found was digoxin, a heart medication derived from foxglove plants. Digoxin has been used for over 200 years to treat heart failure and arrhythmias. But in about 10% of patients, digoxin is converted to an inactive metabolite called dihydrodigoxin by gut bacteria. The bacteria in question are certain strains of E. Lenta. Why do we even have a bacteria in our gut that can break down Digoxin? The leading hypothesis is that gut bacteria maintain these genes to protect their human hosts from plant toxins, and digoxin drug metabolism is just an accidental side effect of an ancient detoxification system.

Digoxin is just one example. Researchers now know of more than 50 drugs that can be directly metabolized by the microbiome. Some examples are L-dopa (for Parkinson’s), certain chemotherapy drugs, antibiotics, and even some antidepressants. The same principle applies: different people have different bacteria with different enzymes, leading to wildly different drug responses.

6. The microbiome might even change people’s behavior

I’ve thought about free will a lot, and written about our somewhat questionable ability to exercise free will before here. It turns out, our microbiome might also be influencing our actions. Gut bacteria produce neurotransmitters like GABA, serotonin, dopamine, and acetylcholine. And these can signal to the brain both through the vagus nerve (a direct neural connection from gut to brain) and through the bloodstream. Over 90% of your body’s serotonin originates in the gut. SCFAs like butyrate and propionate cross the blood-brain barrier and directly modulate brain function. The behavioral implications are still being worked out, but there are some provocative findings. In mice, fecal microbiota transplantation (FMT, or poop transplants) from depressed animals can transfer depressive behaviors to recipients. Dysbiosis correlates with depression and anxiety in humans. In humans, there are similarly spectacular case studies with FMTs. In one famous example, a woman with no prior history of obesity was given an FMT from her overweight daughter, and in the year after reported increased appetite and unintentionally gained over 12kg to become ‘newly obese’.

7. Despite some mind-blowing animal research, translation to humans has been challenging

If you are a mouse, the future of medicine is already here. We can cure almost anything in mice. We can make them thin, fat, anxious, or calm just by tweaking their microbiome. But translating this to humans has been frustratingly difficult.

The problem is that lab mice are biological widgets. They are genetically identical, live in sterile bubbles, and eat exactly the same standardized chow every single day. Their microbiomes are stable and predictable. Humans are the opposite. We are genetically diverse, we are stressed, and we eat chaos. Pizza one day, salad the next afternoon, and then 5 beers and a very unnecessary negroni at night.

Also, the “background noise” of the human microbiome is so loud that introducing a subtle change often gets drowned out. A specific strain might cause a massive weight loss effect in a sterile mouse gut where it has no competition, but when you drop that same strain into a human gut teeming with 100 trillion competitive bacteria, it might just get bullied out of existence before it can do anything.

Then there are the bacteria themselves which are tricky to work with. Many of the most valuable workhorses live in the distal colon, and are strictly anaerobic. The moment they come out of your body, they die within minutes. It is very hard to grow these things in a lab, let alone put them in a pill. Even with the Bifidobacteria and Lactobacillus which we can easily grow and get into the gut intact, to what degree they actually colonize is variable and hard to predict. More than anything, that depends on what you eat. Someone who doesn’t eat any fiber can take all the probiotics in the world, but it won’t have any effect because the bacteria will just starve.

Nevertheless, we’re closing in on more real therapeutic applications.

8. Existing and nascent microbiotic therapies

I mentioned Fecal Microbiome Transplants a while back. The ‘traditional’ way to do this is to literally take Person A’s poo, put in a blender with some salt water, and shoot it up the colon of Person B with an enema. This leaves the biome mostly intact, although the strictly anaerobic bacteria are still going to be dead. This method is really only used therapeutically for one specific killer: Clostridioides difficile (C. diff). C. diff is an opportunistic pathogen that can inhabit your gut harmlessly. But when a strong course of antibiotics wipes out your entire microbiome (the “nuclear apocalypse” I mentioned in the intro), C. diff could be the last man standing. If that happens, it multiplies aggressively and produces toxins that destroy the gut lining. It is often fatal. The irony is that the standard treatment for this infection caused by antibiotics… is more antibiotics. But for recurrent infections, this fails nearly 30% of the time (remember that these people die). FMT, on the other hand, has a cure rate of >90% in these desperate cases.

That’s cool and has saved lives, but there are legitimate concerns and problems with FMT. Is the donor healthy? You can’t know for sure if that person doesn’t have some random pathogen in their biome that’s going to cause other issues. Or that you suddenly get insatiable sugar cravings and become obese, like the woman described earlier. So we would really like to know what specific strains we need to get into the gut, how to get them there in one piece, and how to make sure they grow there.

One of those newly discovered strains is Akkermansia muciniphila, a strictly anaerobic strain that lives deep inside the large intestine. Akkermansia is now known to stimulate GLP-1 (yes, like Ozempic) production through various pathways which slows appetite and regulates metabolism. It is also a Propionate and Butyrate producer. And with some great effort, it can be grown and put into a capsule. This company has created a formulation with several strains of bacteria that in a clinical trial in humans has shown pretty mind blowing effects on metabolism.

Correlation vs causation

The reason I’m excited about the successful supplement trial mentioned above is that in microbiome research it’s been very difficult to determine cause and effect. It seems to be a two-way street. If you eat nothing but hamburgers, you will gain weight and become metabolically unhealthy. That’s correlated with predictable microbiome changes. If, from that state, you start eating healthy the opposite happens, and the microbiome changes again. The biggest driver of microbiome changes is diet, so that makes sense. But then, what’s the point of direct microbiotic interventions? Well, remember that woman again who started eating too many hamburgers after getting a replacement microbiome? Apparently, just editing the biome triggered a behavioral change toward a diet corresponding to that microbiome makeup. What if we gave people with Diabetes Type 2 the microbiome of healthy people? Will they stop eating so many hamburgers? That’s where the GLP-1 stuff comes in. It’s a well documented and rather mind blowing effect of Ozempic, that people just don’t feel like having unhealthy food anymore. It makes them feel physically bad. It seems like that also might be one factors driving the effects of the Akkermansia supplementation.

So it looks like causation runs in both directions. By changing diet, it is possible to change the microbiome, which might ‘lock-in’ that diet to some degree. The biome adjusts to that diet. But it also seems that by changing microbiome, it is possible to change dietary preferences. Many caveats and questions remain here, but it’s certainly exciting.

How to feed your microbiome

Ok, let me just kick in the door here with the key thing I’m almost certain you didn’t know: Some good tasting carbs you thought were unhealthy are actually great!

Rice, pasta and potatoes that have been cooked, and then cooled down

The starch (amylose and amylopectin) in the raw ingredients comes packed in a certain structure, and when you cook them the heat dissolves (gelatinizes) those bonds. Now, in the hot pasta, the starches are kind of ‘floating’ around. When it cools down, they try to pack back together again, but they don’t return to their original state. Instead, they pack even more tightly together through a process called retrogradation. What we end up with are ‘resistant starches’. While the free amylose in hot pasta is broken up into glucose in our small intestine (which can spike blood sugar), the resistant version is not. It passes through the small intestine intact and is instead fermented in our large intestine, producing massive amounts of butyrate. So these cooled down starches are a double win: They escape metabolization to glucose, and they feed good bacteria in the colon to produce butyrate. I knew there was a good reason I’ve always loved pasta salad.

Diversity is key

Remember that different fibers feed different bacteria. You want a diverse ecosystem, not a monoculture. That’s kind of dogma. But why, actually?

Different bacteria produce different SCFAs in different ratios. Butyrate is critical, but propionate (which helps regulate glucose production in the liver) and acetate (which helps regulate appetite and fat storage) matter too. A healthy microbiome produces a mix.

Different bacteria produce different chemicals beyond SCFA production. Some synthesize vitamin K, others make B12. Some transform bile acids to stimulate GLP-1. Some produce specific antimicrobial compounds that keep pathogens in check. Faecalibacterium prausnitzii produces anti-inflammatory compounds. Akkermansia strengthens your gut barrier.

Ecosystem resilience. If you only have one or two species that can produce butyrate and something kills them off (antibiotics, food poisoning, stress), you’re screwed. A diverse ecosystem has functional redundancy - if one butyrate producer goes down, three others can pick up the slack.

Cross-feeding networks. Many bacteria can’t directly process certain fibers, but they can use the breakdown products from other bacteria. Ruminococcus bromii breaks down resistant starch into smaller fragments that it can’t fully use itself, but those fragments feed dozens of downstream species. Without R. bromii, those other bacteria starve even if there’s plenty of resistant starch around².

Different bacteria colonize different locations. Some thrive in the small intestine where there is still oxygen, while others live deep in the anaerobic distal colon. It’s good to have healthy populations everywhere.

The practical application: Eat a wide variety of plant foods. Different fibers recruit different teams:

• Inulin (chicory root, onions, garlic, leeks, asparagus) - Bifidobacterium specialists

• Pectin (apples, carrots, citrus) - Faecalibacterium prausnitzii, major anti-inflammatory producer

• Beta-glucans (oats, barley, mushrooms) - Bacteroides uniformis

• Resistant starch (green bananas, cooled rice/pasta/potatoes, cooked and cooled legumes) - Ruminococcus bromii, the keystone species

• Arabinoxylan (whole grain wheat, wheat bran) - Various Bifidobacterium species

Don’t stress about memorizing which fiber feeds which bacteria. Just eat lots of different plant foods per week. Vegetables, fruits, whole grains, legumes, nuts, seeds, herbs, they all count.

Fermented foods

Yoghurt, kimchi, saurkraut, etc are good because they (can) actually contain bacteria. But the degree in which these bacteria colonize the gut appears very low. That means the bacteria in these foods are more like tourists that pass through. They can still do beneficial things during their short visit. Studies show people who regularly eat fermented foods have lower inflammation markers and more diverse microbiomes. In one Stanford study, a high fermented food diet decreased 19 inflammatory markers in just 10 weeks.

Critical note about pasteurization: Most commercial sauerkraut, pickles, and other fermented vegetables you find on regular grocery store shelves are pasteurized. That process kills all the bacteria. You’re basically eating pickled cabbage with none of the probiotic benefits (but still prebiotic benefits though). Look for unpasteurized or “raw” versions in the refrigerated section. Real fermented foods should say “live cultures” or “unpasteurized” on the label, and they need to be kept cold. If it’s shelf-stable in the condiment aisle, the bacteria are dead.

For yogurt and kefir, this is less of an issue since they’re almost always sold refrigerated with live cultures.

Polyphenols feed the right bacteria

Polyphenols are compounds in colorful plant foods that our bodies can’t digest well, but our gut bacteria love them. Foods high in polyphenols:

• Berries (especially blueberries, blackberries)

• Dark chocolate (again, I knew it!)

• Green tea

• Coffee

• Red wine (in moderation)

• Olive oil

• Nuts

Akkermansia municiphila, that GLP-1 stimulating bacteria I mentioned, specifically thrives on polyphenols. It also feeds on the mucin layer of your gut lining, which gets replenished partly in response to polyphenol consumption.

In practice, what I’m doing after this study

• I’m trying out the Pendulum Probiotic supplement, that was proven to colonize and increase Akkermansia in a clinical study (I’m not sponsored, I wish I was).

• Looking at the list of polyphenol containing foods, I think I’m in good shape, but I’m definitely going to eat more dark chocolate again.

• Increasing my intake of easy to digest fiber: resistant starches and soluble fibers. Apples, berries, cooled starches, bananas. I had been avoiding several things in this category, because I thought they were mostly sugar. Turns out they’re actually great!

• Somewhat moderating my intake of raw fiber. I had been eating tons of raw kale and spinach, raw nuts, etc. Basically, everything as raw as possible. Turns out that’s a mistake. Cooking destroys some nutrients like Vitamin C, but it makes others much easier to extract. Lycopene in tomatoes increases 2-3x when cooked, Your body gets a lot more nutrients out of kale if it’s cooked. Gotta mix it up and eat cooked veggies too.

The key eye-opener for me is that I had misinterpreted the timeless advice “eat food, mostly plants, not too much” in two key ways: Firstly, by somehow assuming that it’s always better when those plants are raw, and secondly by putting rice, pasta and bread into the ‘unhealthy carbs’ bucket mentally instead of more accurately considering them plants as well.

Hope you found this interesting and educational. Personally felt it was a nice break from the AI and tech stuff. If you have any comments or experiences to share, please reply or comment, I’d love to hear them.

Sources and factuality

Between 2005 and 2010 I studied Biomedical Engineering. While I forgot a lot, and science has obviously progressed, core metabolic and cell biology concepts were drilled in so thoroughly that I was able to draw on this as scaffolding to understand the subject.

Throughout the writing above, I’ve tried to source every key claim to solid, peer-reviewed scientific literature (just click the links, or not, at least they’re there).

This is the first Substack post I’ve ever written with significant assistance from AI, mostly Claude. It not only helped fact-check, surface sources, and give feedback on the structure and text, but I have also actually incorporated passages written by Claude directly for the first time (see if you can spot them). Nevertheless, it goes without saying that I’ve double-checked all sources and >90% of the words here were not produced by energy hungry Nvidia chips but by some good old glucose oxidation and fiber fermentation powered brain and muscle cells.

1. Using ATP, or in some bacteria the closely related GTP as the energy ‘currency’ is a core fundamental property of life. It is universal to ALL known lifeforms. It is thought to have emerged billions of years ago with the very first living cells. Things happening in the body are often explained in chemical terms, but you can also look at them in mechanical terms. There really are all kinds of tiny machines inside your cells. These machines work by physically changing shape through chemical changes propelled along by the ATP cycle.

2. Everyone who studies biology wonders at some point why nature has such long signalling pathways. Some molecule gets produced and the only function it has is to pass on a signal to another place. Why not just do the final thing, instead of taking multiple intermediate steps? We don’t really know the answer. It’s just evolution probably. These pathways were formed over hundred of millions of years. It’s like technical debt in nature. Nobody really knows exactly how it works anymore, but it does, so let’s not touch it!