Lately I’ve been turning my attention back to one of the original lines of inquiry here on this blog — the question of what effect high dose Resistant Starch (mostly in the form of raw potato starch) is having on the gut bacterial profile in individuals. I think it’s an important question for a few reasons. Despite the many reports of positive effects gained from this experiment — digestive, sleep, mood, metabolic, etc. — I still believe it’s necessary to keep some healthy skepticism. After all, it isn’t like we’ve cured cancer just yet. Having a bunch of people who have normally been on a low fiber Western diet all of a sudden throw in high doses of one very specific type of fermentable fiber is a pretty novel thing. And there are many examples of this or that protocol having great effects at the outset, only for things to level off, or even go south, when practiced chronically for a prolonged period (vegan diet, anyone?). But most importantly, I simply have the nagging feeling that there are much greater gains to be made from finding out what is going on and what should optimally be going on.
What initially spiked my skepticism was the fact that in the one gut sequencing report we have of an individual “high dosing” on RS, the results seemed…not quite what we would expect. (By the way: I am DYING for some new data from other individuals. Hopefully we’ll seem some trickle in soon.) In these two initial posts, we wondered why Tim’s results showed higher than average Bacteroidetes, and lower than average Firmicutes, specifically from the the two important Clostridia clusters Lachnospiraceae and Ruminococcaceae. These are the plant fiber degraders and butyrate producers, and have a strong association with gut health. Bacteroidetes, on the other hand, are associated with a lower-fermenting gut, typical of the standard Western diet and its low fiber content. American Gut founder Jeff Leach and writer Michael Pollan — both enthusiastic plant eaters — had far more of these (especially Jeff). But even when compared to the average American, Tim had less of these, and more Bacteroidetes.
Given that one of RS’s purported benefits was the stimulation of butyrate production, it seemed odd that Tim’s butyrate-producing bacteria would be lower than the average person’s. Specifically, his Lachnospiraeceae — which contain major butyrate producers like Roseburia and Eubacterium — clocked in at 4.5%, last in his “most abundant.” I’ve seen a few other AmGut reports of non RS supplementers where this group registers at the top of the list at around 40%. Heck, even Tim’s wife’s report showed slightly more of these bacteria, and I don’t think she was supplementing with RS at all at the time:
(By the way: Tim and I speak often and he has always been very encouraging of the use and analysis of these reports. Many thanks to him for giving us something to work with.)
Well, in the end, we concluded that RS was probably strongly stimulating a very select number of species within the butyrate-producing Clostridia groups — ones shown to have good starch degrading capability in vitro. And we left it at that.
But I’m now beginning to question whether that’s the whole story.
Since those posts, we discussed here an interesting experiment done by Leach showing pretty conclusively that Firmicute growth was a direct result of consuming fermentable fibers from plants, and that in the absence of that consumption, Bacteroidetes would rise. Below you can see the dramatic reversal.
This was presaged in a comment by Leach in a previous article:
In addition, as pH shifts away from acidic, the genus Bacteroides can also bloom as well, gaining an ecological niche in this less acidic environment courtesy of a low carb diet. For those of you keeping score, many talk about the American gut in general being dominated by Bacteroides as a function of our high fat, high sugar diet. The reality is, it might have to do with what we are not eating – dietary fiber (of all kinds). The all-important butyrate producers Roseburia spp. and Eubacterium also drop in abundance as pH shifts away from acidic as well.
So from this, we took away the idea that the acidity produced by fermentaton is the driving force and a useful, “unifying” theory of gut health. A rise in acidity favors the Firmicutes and Actinobacteria (good guys which include the Bifidobacteria), and a drop favors Bacteroidetes and the dreaded Proteobacteria. And we speculated that Bacteroidetes dominance may not be such a great thing — Bacteroidetes are all gram-negative endotoxin producers, and studies have made numerous correlations to disease (Type 1 Diabetes, IBD, Obesity, Metabolic Disease). And they seem to rise when you smoke. To be clear, they are also quite clearly human commensals and not outright pathogens (though certain species do seem to turn pathogenic in certain circumstances). The endotoxins they produce are vastly less toxic than those of Proteobacteria (though that is balanced by the fact that they vastly outnumber Proteobacteria). So it’s complicated. But the more important point is that I have yet to see any disease correlations with the Clostridia clusters who, again, are the major butyrate producers and considered by many to be a health marker of the microbiome. It is this group — along with Actinobacteria (ie, Bifidobacteria) — that seems to be the marker of fermentation, proper acidity, and gut health. And we have plenty of correlations showing disease when they are in low number.
Oh, and just in case this is all a little too N=1 for you, there was another recent study done that showed this exact same effect. In fact you may be familiar with it, as it got a decent amount of press and was featured in an NPR article. The study split people into two groups. One ate a completely animal based diet (meat, cheese). The other ate a completely plant-based diet. The results:
The animal-based diet increased the abundance of bile-tolerant microorganisms (Alistipes, Bilophila and Bacteroides) and decreased the levels of Firmicutes that metabolize dietary plant polysaccharides (Roseburia, Eubacterium rectale and Ruminococcus bromii).
As you can see, the same effect. (The annoying part of this study and the media reports about it was that they predictably drew the conclusion that meat and fat caused the effect. While there may be some contribution from the animal foods, the fact that it was a “bacteria starvation” diet is far more relevant. As Leach’s experiment showed, it is more likely that it was a complete lack of plant fiber, not the meat and fat, that caused the “bile-loving” bacteroides to grow in the animal-based diet. It’s about fermentation, or the lack thereof, and the resulting pH. Why didn’t they have a meat + plant group? Because it would’ve interfered with their simple and headline-grabbing conclusion?)
So when we see lower than average butyrate-producing Clostridia and higher than average Bacteroidetes in an RS supplementer’s gut report, there are two basic facts that should give us pause when considering the purported benefits of RS supplementation:
1) The abundance of Clostridia has been found to have a direct and positive correlation with butyrate production. This is something that definitely raises questions about the value of high dose RS supplementation. The more Clostridia, the more butyrate. And vice versa.
2) The colonic pH that is conducive to Bacteroidetes has been shown to be incompatible with a high fermentation, butyrate-producing gut. It is a pH that inhibits active fermentation and butyrate production from butyrate producers like Eubacterium and Roseburia, and instead favors Bacteroides growth and its metabolites. This, too, raises doubts about the benefits of high dose RS supplementation.
In other words, if RS supplementation results in higher than average Bacteroidetes and lower than average Clostridia, it is likely not delivering on its promise.
Or at least half of its promise.
You see, taking all of this into account, we have just one problem — Tim also had a superhuman count of Bifidobacteria, and no Proteobacteria whatsoever – the kind of thing you’d expect to see in a healthy, high fermentation gut. Pretty uncommon, and unlikely to be just a fluke. Pretty clearly the result of some outlier factor in Tim’s diet. In other words, it’s quite clearly the potato starch. The RS is doing something right. How do we explain this?
One of the foundational concepts underlying gut microbial dynamics and microbiome diversity is substrate competition. “Substrate” is just a fancy word for any food that bacteria feed on. Substrate competition refers to the idea that the outcome of what you feed your gut — what bacteria grow and what byproducts (short-chain fatty acids like butyrate) are produced — is largely decided by the competitive dynamic between bacteria in the gut.
Different bacteria are adapted to metabolize different types of plant fibers. And even if one type of bacteria is able to break down a particular type of fiber, it doesn’t mean it will. Remember, it exists in a large, competitive community, and there may be other bacteria that are much more suited to metabolize that particular type of fiber. In other words, stimulating the growth of specific groups of bacteria relies on the consumption of fiber that they can preferentially feed on — fibers that they have carved out a specific niche for degrading.
That’s why in vitro experiments looking at how well a specific type of bacteria grows on some type of fiber, and what it produces as a result, aren’t very useful. That’s sort of like watching a baseball player swing a bat really hard in training, and then predicting that he and his team are going to win the World Series.
As it turns out, starch is a pretty interesting substrate. It’s highly bifidogenic, that we know from the research, and Tim’s results show that pretty clearly. In fact I’m betting this is going to be the most common result we see in the gut reports from RS experimenters: crazy high bifido’s. And this is why: bifidobacteria are known to have a unique capability whereby they adhere to resistant starch granules, allowing them to degrade RS. It is this capability that allows bifidobacteria to compete quite well for this substrate, and this is very likely why RS has such a bifidogenic effect and why we see that in Tim’s report. (And as we’ve discussed before here and here, the degradation of RS by bifidobacteria is also crucial for its fermentation by the butyrate-producers Roseburia and Eubacterium. Cross-feeding: they seem to need bifido’s to “pre-digest” the RS, and they also feed off of the acetate and lactate that the bifido’s produce when they degrade RS.)
But starch is also preferred by one other group of bacteria. A group that has also carved out a specific niche for starch degradation.
That would be the Bacteroides.
The Bacteroides (which are the predominant genus within the Bacteroidetes phyla and are interchangeable for our present purposes) seem to have quite the sweet tooth for starch.
Below, from a brand new, page-turning, thrilling read on Resistant Starch, we find a comparison of starch-degrading ability between Bacteroides, Actinobacteria, and Firmicutes:
As you can see, Bacteroides seem to come out on top. More Bacteroides strains show starch-degrading ability than any other phylum. Actinobacteria — the Bifidobacteria — come in at second. We expected them to do well on starch. Firmicutes: last place. This table is based on some pretty old research (1977), and the analysis allows for the fact that Firmicutes may have been neglected in culturing when this study was done, so let’s keep going.
Well look at this. Looks like we’re late to the game. From a 1989 study on B. thetaiotamicron, the leading Bacteroides commensal in the human gut:
These findings indicate that starch utilization by B. thetaiotaomicron apparently does not involve secretion of extracellular enzymes. Rather, binding of the starch molecule to the cell surface appears to be a first step to passing the molecule through the outer membrane and into the periplasmic space.
It looks like Bacteroides can do something quite similar to what the Bifido’s do: bind starch, hold on tight, and munch away:
The intestinal symbiont Bacteroides thetaiotaomicron uses five outer membrane proteins to bind and degrade starch.
These starch utilization systems are presumed to help Bacteroides effectively compete for starch:
It appears that an elaborate system of starch-binding proteins and periplasmic hydrolases enables this bacterium to sequester and degrade starch molecules, presumably allowing it to compete more effectively for the available substrate.
And it seems that it isn’t just B. thetaiotamicron. These starch utilization systems seem to be shared across the Bacteroides genus. Other leading species, like B. ovatus, B. vulgatus, and B. fragilis have been found to possess the same capability.
And in one study where this competition was tested, we see how things might actually play out:
Although able to use amylopectin in pure culture, Roseburia sp. strain A2-194R competed poorly for this substrate in the mixed community.
The results of our FISH analyses show that the fermentor conditions favored the growth of Bacteroides strains; the proportion of these organisms increased from 10% in the fecal inoculum to 40 to 60% of the total bacterial count, while the proportion of gram-positive anaerobes fell.
And finally, this study wraps it all up nicely with a bow on top:
Non-adherent Bacteroides sp. and Bifidobacterium sp. have been shown to outcompete gram-positive bacteria (such as Firmicutes) for easily hydrolysable starch, while Lachnospiraceae and Ruminococcaceae persist in fibrolytic communities and are uniquely suited to degrade a wide variety of recalcitrant substrates.
Is it conceivable that starch, in a particular intestinal environment, would favor Bacteroides growth? Perhaps, in certain conditions, Bacteroides might be able to compete effectively for starch? Maybe:
Dietary intake can also affect the pH in the proximal colon, and this appears likely to be a key factor determining butyrate production. In an in vitro fermentor study with a faecal inoculum, it was found that the two major butyrate- producing bacterial groups, Roseburia/E. rectale species and F. prausnitzii, thrived at pH 5.5, whereas their population declined dramatically at pH 6.5, with Bacteroides spp. becoming dominant. In accordance with the population changes, butyrate was the main fermentation product at pH 5.5, while acetate and propionate became the main products at pH 6.5.
Ah yes, pH again. Not so optimal pH = Lots of Bacteroides = not that much butyrate. So it seems that what else is going on down there is pretty important — crucial — to how this might all play out. We’ll get back to that.
The interesting thing about Bacteroides, of course, is that while it seems they like starch, they don’t really need it. In fact, they don’t need you to feed them anything. How come? Because they’re very happy to feed on you. Your mucins, specifically:
Mucin-type O-glycans are the primary constituents of mucins that are expressed on various mucosal sites of the body, especially the bacteria-laden intestinal tract. Mucins are the main components of mucus, which is secreted by goblet cells and forms a protective homeostatic barrier between the resident microbiota and the underlying immune cells in the colon.
And Bacteroides love those O-glycans. Feel like starving them? That’s fine:
According to nutrient availability, B. thetaiotaomicron can redirect its metabolism from dietary polysaccharides to host-derived polysaccharides, including mucus, and vice versa, and further refine its niche specificity. This ability of B. thetaiotaomicron to grow on mucus contributes to its colonization and persistence in the GIT.
This study found that Bacteroides “revealed a capacity to turn to host mucus glycans when polysaccharides were absent from the diet.” Some might call that opportunistic. Me? Enterprising.
As Bacteroides species are also potent mucin degraders, one may therefore assume that mucin degradation may be a more important process in the transverse colon than in the ascending colon.
Might this mucin degradation have something to do with the Bacteroides correlation to inflammatory disease? Maybe:
Another possible mechanism by which members of the Bacteroidetes probably contribute to the pathogenesis of colitis is by the production of mucin- degrading sulphatases. Elevated levels of bacterial mucin- desulphating sulphatases have been reported for patients suffering from active UC. Furthermore, the existence of enzymes that will partially desulphate mucins has been demonstrated for B. thetaiotaomicron, B. fragilis and for Prevotella spp., suggesting that members of the Bacteroidetes could contribute to chronic inflammation by an impairment of the barrier function of the epithelial cell layer.
Now do you see why blaming the rise of Bacteroides on the consumption of animal-derived foods seems a bit…off? Bacteroides don’t need that fancy grass-fed ribeye you just scarfed down. They have you and your host-derived glycans. They love it when you starve your gut, because they have a plan B, and a lot of the other guys don’t.
As such, Bacteroides are hardy bacteria. They are survivors. It’s almost as if they are our “default” gut inhabitants. In this Wired piece, we see that 9 months after a subject is treated with antibiotics, their entire gut is populated by just one bacteria: B. thetaiotamicron. Creepy.
Alright. Time to stand back and take a deep breath before we descend into mass hysteria.
Here’s the point of all this: I think it’s clear that putting a ton of emphasis on one, sole source of fermentable fiber doesn’t make much sense for your microbiome, and never did. Why? Because the microbiome is a highly complicated ecosystem, and the idea that one can game it this way is unrealistic.
Does this mean RS intake, in general, is somehow counterproductive? Of course not. Could singling it out as a sole or predominant source of fiber with a Western diet-shaped gut perhaps induce some sort of non-ideal imbalance? Maybe. But it’s a major ancestral source of prebiotic fiber, without a doubt. And there is plenty of good research showing that RS does in fact do what it’s supposed to. This study tested starch competition between a few butyrate producing Clostridia and Bacteroides, and the butyrate producers came out on top. And here we have three in vivo human studies showing the same — resistant starch significantly boosts Clostridia and butyrate in human subjects. That’s pretty good evidence, and shows that our initial conclusion about how RS is working wasn’t off base. It was probably right. In theory, that is.
But in reality, things are more complicated. As Tim’s gut report shows, results are probably going to be highly variable. It could very well be that Tim’s gut report is a total outlier. But I sort of doubt it. I think what it probably means is that results will be all over the place, and that the gut you start out with will be a critical factor. And probably overall diet. Which brings us to our final and concluding point.
In the end, it shouldn’t really matter how one single source of fiber affects one’s microbiome, because that’s not what your gut was built for — or at least shouldn’t be. Luckily, there is one great equalizer in all this: diversity. Plant diversity. Microbial diversity. A diverse intake of fermentable fiber that feeds a diverse population of bacteria.
As we said before, substrate competition means different bacteria come to the party for different reasons. Amylolytic (ie, starch-degrading) bacteria are just one subset of the microbiome. Remember how the overall pH of the gut — determined by fermentation activity — is crucial to allowing starch-degrading butyrate producers to actively ferment? How do you expect that to happen if you only invite a few bugs — no matter how important they might be — to the party? It seems that a more diverse guest list may just tip the balance:
Since Bacteroides spp. appear to be less tolerant of growth inhibition by weak acids at reduced pH than many gram-positive species, it appears that the mildly acidic pH creates the opportunity for the gram-positive species to compete successfully with Bacteroides for the polysaccharide substrates. The consequence of these differences in pH tolerance, via their effects on the balance of the microbial community, was a four-fold higher butyrate concentration in the fermentor at pH 5.5, compared to that at pH 6.5.
But it goes even deeper than that. There’s another defining principle of gut microbial dynamics, and that’s cross-feeding. There is a mind-boggling level of interdependence down there. Did you know that the acetate formed by other bacteria is a major contributor to butyrate production from bacteria like Roseburia and Eubacterium? Much of the butyrate produced is not from direct feeding, but by feeding on acetate (an SCFA) produced by the feeding of other bacteria. And that’s probably just one of a gazillion co-feeding interactions. In other words, cross-feeding is another reason to invite everyone to the party: they all need each other, or it doesn’t work. And until that’s all figured out (and I don’t think that will happen any time soon), doesn’t it make sense to throw a fiber party down there and let the biology sort the rest out?
But hey, don’t take my word for it. People who have devoted their entire careers to this stuff seem to have come to that simple conclusion:
“Eat as many high-fiber fruits and vegetables and legumes as you can,” says Stanford University’s Justin Sonnenburg, a microbiologist who studies how diet impacts bacteria in the gut. “Our hypothesis is that a variety of plant fibers supports a diversity of gut microbes.”
And if there’s one constant, non-controversial marker of gut health that I’ve come across consistently throughout the research, more than any other, it’s diversity. Likely because a diverse gut is a resilient one that is firing on all cylinders, fermenting anything and everything you throw at it — which means a healthy, low pH environment — and because diversity supports a robust, complex cross-feeding community. And that’s probably just the tip of the iceberg.
Is RS a panacea? Nah. That’s a pretty strong word. But it’s an important slice of the pie. In future posts, we’ll look at the evidence for what the rest of that pie might actually look like.