The last decade has brought enormous advances in our understanding of the microbial communities that call our gut home. It turns out that what we eat has a substantial effect on the gut microbiome. Read on for a summary of some of the most influential papers in this exciting area of research, and how they offer new insight into nutrition science.
Once considered the “forgotten” organ, we now know that the gut microbiota regulates many aspects of our health and physiology. Thanks to a combination of controlled animal studies, human dietary interventions, and advanced sequencing technologies, diet has emerged as a key factor in determining gut microbiota composition and function.
Our gut ecosystem is surprisingly malleable, dependent on both our short-term and long-term food intake and the quality and quantity of various food components. In turn, microbes also influence our digestion and absorption, caloric intake, micronutrient availability, and even our blood glucose response to foods. More than ever, we are learning that “you are what you eat” may more accurately be “you are what your microbes metabolize”.
For this first article in this diet-microbiome series, I will review some of the most seminal papers that have given us a new understanding of human nutrition. In part 2, I’ll discuss the research for today’s popular diets, and what we know about how they affect the gut microbiome.
Long-term dietary patterns shape gut microbiota composition
Published in 2010, a comparative study of children from urban areas in Europe and children from a rural African village in Burkina Faso was among the first to link long-term dietary patterns with the composition of the fecal microbiota.1 Children from Burkina Faso had a significant enrichment of the phylum Bacteroidetes – particularly Xylanibacter and Prevotella, which likely assist in the degradation of cellulose, xylans, and other complex polysaccharides in their high-fiber diet. Meanwhile, European children had higher levels of Enterobacteriaceae, a family of bacteria that has been associated with several chronic inflammatory diseases and reflects their Western diet much lower in fiber. High levels of Enterobacteriaceae are now considered a microbial signature of dysbiosis.
Short-term dietary changes rapidly alter the gut microbiota
Short-term dietary influences are also known to affect gut microbial diversity and activity. A seminal paper by David et al. (2014) found that dietary changes alter gut microbiota composition and gene expression within as little as 48 hours.44 Switching individuals from their a high-carbohydrate, plant-based diet to a ketogenic, animal-based diet significantly increased the diversity of the gut microbiota within two days, a feature that was lost within two days of returning to their normal diet. In other words, the gut microbiota has a high degree of “metabolic flexibility”, being able to rapidly shift to metabolize the foods available to the host.
Diet particularly impacts the gut metabolome
However, studies in humans using less extreme differences in diet suggest that changes in gut bacterial communities may be much more modest. A study by Wu et al. (2016) compared the fecal microbiota and metabolome (the sum total of all of the metabolites) of healthy vegans and omnivores living in an urban Western environment.45 They found surprisingly modest differences in the composition of the gut microbiota using 16S rRNA gene sequencing, but large variation in the metabolome. Overall, their results point to a greater role for diet as a modulator of the bacterial metabolome or compositional changes at the species or strain level than as a factor that directly regulates gut microbial composition at higher taxonomic levels.
Gut enterotypes: useful or not?
A 2011 study led by the laboratory of Dr. Peer Bork in Germany identified three distinct compositions, or enterotypes, of the human gut microbiota.2 Each enterotype is dominated by a different genus of bacteria – Bacteroides, Prevotella, or Ruminococcus – which co-occur with various other genera. In other words, these enterotypes represent three ecologically distinct stable states of the gut microbiota. Further research published later that year by researchers at the UPenn Perelman School of Medicine suggested that enterotypes were strongly associated with long-term diets, with Bacteroides associated with higher protein and animal fat intake and Prevotella associated with higher amounts of carbohydrates.3
However, more recent research has challenged the concept of enterotypes, suggesting that they may be more of a blurred gradient than distinct compositional clusters. Indeed, the same study above found that the Ruminocccus enterotype was partly merged with the Bacteroides enterotype. Moreover, a controlled 10-day dietary intervention was insufficient to get an individual to switch enterotypes entirely.3
Microbes ferment dietary fiber into short-chain fatty acids
One dietary component that seems to have a particularly influential effect on the gut microbiota is fiber. The consumption of dietary fiber has been shown to have beneficial impacts on food intake, body weight, and glucose homeostasis, and numerous epidemiological studies have found associations between higher fiber intake and a reduced risk of diabetes, cardiovascular disease, irritable bowel syndrome, inflammatory bowel disease, and colorectal cancer.48 Certain types of fiber are considered prebiotic – meaning that they are selectively utilized by host microbes and confer a health benefit.4
In 1987, years before the microbiome field was even borne, researchers observed the fermentation of dietary carbohydrates in the human large intestine into short-chain fatty acids (SCFAs).5 While the mammalian digestive tract lacks the enzymes necessary to metabolize complex polysaccharides, our microbes have these enzymes and produce SCFAs as a byproduct. We now know that these small metabolites have large and far-reaching effects on host physiology.51
For more on short-chain fatty acids, see my five-part article series.
A Western diet low in fiber drives inflammatory disease
The modern Western diet contains large amounts of highly processed, refined foods and much less fiber and vegetables than the diets of developing countries or traditional cultures in the Western world even 40-50 years ago. The average daily fiber intake among humans in Westernized countries is approximately 20-25 grams, compared to agrarian or hunter-gatherer societies, which are estimated to consume up to 60 grams of fiber per day.6 Using humanized mice — germ-free mice that are colonized with human fecal material – Turnbaugh et al. (2009) found that a Western diet confers rapid deleterious changes on the gut microbiome.7 After just a single day on the Western diet, there were significant alterations in composition, metabolic pathways, and gene expression. These microbial changes preceded weight gain and adiposity.
Diet-induced extinctions compound over generations
In 2016, Sonnenberg et al. further demonstrated the importance of dietary “microbiota-accessible carbohydrates” (MAC) for supporting the health of the gut microbiota. 50 Switching humanized mice from a high-MAC diet to a low-MAC diet reduced microbial diversity and drove certain taxa to low abundance. This was largely reversible in a single generation when the mice were put back on the high-MAC diet. However, when a low-MAC diet was maintained over multiple generations, taxa driven to low abundance were not efficiently transferred to the next generation. After four generations, even a high-MAC diet could not restore gut microbial diversity. Considering the recent shift in the modern diet towards processed, low-MAC foods, these findings could explain the lack of diversity in the Western microbiota compared to hunter-gatherer societies.
Could dietary plant diversity be key?
A study published in mSystems by Dr. Rob Knight’s lab mining data from the American Gut Project found that regardless of self-reported dietary category (vegetarian, vegan, or omnivore), individuals who consumed more than 30 different types of plants each week had higher microbial alpha-diversity than those who consumed less than 10 plant foods per week.8 The beneficial SCFA-producers Faecalibacterium prausnitzii and Oscillospira were among the genera significantly more abundant in the fecal microbiota of people consuming a wide variety of plant foods. Looking more closely at the stool metabolome, they found that dietary plant diversity also led to increased production of conjugated linoleic acid (CLA), likely reflecting a greater abundance of linoleic acid-to-CLA converting Lactobacillus plantarum and Bifidobacterium.
I’ll have more to say on fiber and plant diversity in part 2 of this series!
Cooking food impacts the gut microbiota
Most recently, Carmody et al. (2019) elucidated the differential effects of cooked and raw food on the gut microbiome of both mice and humans.28 Surprisingly, there were no significant differences in the gut microbes of mice fed raw vs. cooked meat. However, raw tubers, which are high in indigestible starch, led to lower alpha diversity than cooked tubers. Through a series of follow-up experiments that tested raw and cooked carrots, beets, peas, corn, and white potatoes, the research group determined that starch digestibility was a key factor in shaping the gut microbiota response to cooking. Notably, consumption of raw tubers led to increased indices of gut microbial cell damage and decreases in bacterial load and cell activity that were not observed in the groups consuming cooked tubers. To evaluate the effects in humans, they fed healthy participants matched raw or cooked plant-based diets over two 3-day periods in a randomized cross-over design. Consumption of cooked plants resulted in higher microbial beta diversity. Evidently, the adoption of habitual cooking has played a major role in shaping our gut microbiomes into what they are today.
An obesity-associated microbiome harvests more energy from the diet
Our understanding of the connections between diet and the gut microbiota really started with animal studies of obesity in Dr. Jeff Gordon’s lab in 2006. These early studies found that the microbiota of obese mice is enriched in genes that allow for increased energy harvest from the diet.9 Transplanting the gut microbiota of these obese mice into lean, germ-free mice transferred the increased capacity for energy harvest and the obese phenotype, without an increase in food consumption.
Further research by the same group in 2008 showed that diet-induced obesity induces “marked but reversible alterations” in the distal gut microbiome of mice, and particularly increased a class of bacteria within the phylum Firmicutes.10 A 2009 study from Gary Wu’s lab corroborated these findings and found even larger changes in the gut microbiota upon switching to an Western diet, including a decrease in Bacteroidetes and an increase in both Firmicutes and Proteobacteria.11 This was true even in mouse strains that were genetically resistant to obesity.
The full extent of obesity-microbiome connections are outside the scope of this review but will be a topic I cover in depth in a future article.
Seasonal cycling in the gut microbiome of hunter-gatherers reflects changes in diet
A study published by the Sonnenberg lab in 2017 looked at the effects of seasonality in the Hadza hunter-gatherers in western Tanzania.12 Collecting sets of longitudinal samples for more than a year, they found striking differences in the gut microbial communities between seasons, with some taxa dropping to undetectable levels in one season, only to flourish in the next. Diversity peaked in the dry season, when the Hadza eat more meat but rely on Prevotellaceae to break down the complex fibers in tubers and fruit from the baobab tree. In the wet season, the Hadza eat more honey and berries, and Prevotellaceae populations collapsed. Most intriguingly, the microbes that are most volatile between seasons in the hunter-gatherers are rare or entirely absent in industrialized populations.
A fasting-mimicking diet boosts protective gut microbes and promotes intestinal regeneration
Recent research from Dr. Valter Longo’s lab suggests that a fasting-mimicking diet (FMD) can stimulate the growth of protective gut microbes and promote gut regeneration in inflammatory bowel disease (IBD). These cyclic low-protein, very low-calorie diets, followed by a return to a normal diet, increased the abundance of protective microbial taxa, including Bifidobacteriaceae and Lactobacillaceae, which bloomed after two days of re-feeding. In contrast, water-only fasting increased levels of Paraprevotellaceae and only partially attenuated symptoms. Fecal transplant of an FMD-treated microbiota into germ-free mice conferred protection against colitis, suggesting that the changes in microbial communities were responsible for the gut-protective effects of the FMD.
The gut microbiota mediates the anti-seizure effects of the ketogenic diet
A study published in the journal Cell in May 2018 found that the beneficial effects of the ketogenic diet on epileptic seizures are also mediated through the gut microbiome.13 The ketogenic diet increased the abundance of the beneficial mucus-associated microbe Akkermansia muciniphila and Parabacteroides spp., and favored bacterial pathways that increased the ratio of the inhibitory neurotransmitter GABA to the excitatory neurotransmitter glutamate in the brain. Treating the mice with broad-spectrum antibiotics abolished the protection against seizures, suggesting that the changes in the gut microbiota were necessary for the therapeutic effect of the diet.
For more on this, see my previous article on how high-fat and ketogenic diets affect the gut.
Diet and archaea, fungi, and viruses
What about non-bacterial members of the gut microbiome? A few studies have begun to look at the effects of diet on the human archaeome, mycobiome, and virome. Hoffmann et al. (2013) looked cross-sectionally at the gut microbiota of 98 individuals that conducted detailed dietary inventories.14 They found that diets high in carbohydrate were associated with a higher relative abundance of the methanogenic archaea Methanobrevibacter and the yeast Candida. These same microbes were negatively associated with diets high in protein and fat. Minot et al. found that the human virome can also change following alterations in dietary sugar, fat, and fiber content.15
Our co-evolution with microbes was (and still is) strongly influenced by diet
Several studies have elucidated the role that diet likely played in the co-evolution of our gut microbiome. Surveying the fecal microbiota of humans and 59 other mammalian species, a 2008 study by the Gordon lab found that bacterial communities “co-diversified with their hosts”, and that the composition of the gut microbiota clustered more according to the degree of dietary similarity than according to the proportion of shared genetics.16 A follow-up metagenomic analysis of 33 mammalian species in 2011 corroborated the importance of diet in differentiating gut microbial communities and their functions.17 Other studies have provided more recent examples of the impact of host diet on the evolution of the gut microbiota. A 2010 study provided strong evidence for the transfer of carbohydrate-degrading enzymes from seaweed-colonizing marine bacteria to the gut microbiota of Japanese individuals!18
Stay tuned for more about the co-evolution concept and how this might shape our understanding of the optimal diet for gut health in part 2.
Diet trumps genes in shaping an individual gut microbiome
Within an individual organism, diet plays a greater role in shaping the gut microbiome. A study published by the Turnbaugh lab in 2015 looked at the effects of dietary perturbations with a high-fat, high-sugar diet across five inbred mouse strains, four transgenic lines, and an outbred wild-type mouse line.19 Regardless of mouse genotype, a high-fat, high-sugar diet produced largely consistent shifts in the microbial communities, increasing the relative abundance of Firmicutes and decreasing the abundance of Bacteroidetes. These results that the effects of dietary intake overshadow any preexisting differences due to host genotype. Other studies have similarly suggested that diet and lifestyle are likely the largest contributors to the inter-individual variation in the composition of the human gut microbiome.20
Past diet exposures influence gut microbiota response
Turnbaugh et al. (2015) went on to demonstrate that the structure and function of the gut microbiota are also, at least in part, influenced by prior dietary history.19 This is an example of a phenomenon known as hysteresis, where the state of a complex system depends not only on its current environment but also on past exposures. In this case, oscillating mice between a normal chow diet and a high-fat, high-sugar diet (a.k.a. “yo-yo dieting”) revealed a number of taxa that had a magnified response to later exposures.
This finding was corroborated by Thaiss et al. in 2016 using a similar mouse model of recurrent obesity. They found that mice who had been put on a high-fat, high-sugar diet in the past gained more weight when re-exposed to the same diet later in life. Notably, antibiotic treatment of the mice abolished the metabolic memory. Metabolomic analysis found that mice that had been previously obese seemed to have low levels of the flavonoids apigenin and naringenin, and supplementing the diet of mice with these flavonoids reduced future weight gain.
Gut microbes influence nutrient bioavailability
The gut microbiota also impacts nutrient bioavailability. A seminal study from the lab of Dr. Federico Rey in 2017 identified human small intestinal bacteria that can compete with the host for nutrients.21 Using gnotobiotic mice colonized with a specific, known community of microbes, they demonstrated that certain strains of E. coli took enough choline from the diet to result in DNA methylation patterns and disease manifestations that mimicked overt choline deficiency. Other dietary components may interact with the gut microbiota in a more positive way. Bess et al. (2019) demonstrated the co-occurrence of multiple genes required for the conversion of plant-derived lignans into enterolignans.22 These studies emphasize the importance of considering microbial networks in host-microbe or host-metabolite interactions.
Microbiota-directed complementary foods can ameliorate childhood undernutrition
Most incredibly, the gut microbiota also has potential to treat undernutrition. Dr. Gordon’s latest work has identified certain strains of bacteria that are present in healthy children but absent in children with severe acute malnutrition in Bangladesh.23 These “growth-discriminatory strains”, including Faecalibacterium prausnitzii, Ruminocococus gnavus, and several Clostridium species were able to ameliorate the malnourished phenotype in a humanized mouse model. The researchers were further able to identify microbiota-directed complementary foods that can selectively stimulate the growth of these importance species! In both animal studies and human feeding trials, they have shown that these unique prebiotics can improve gut barrier function, bone growth, brain development, and immune function.
Responses to prebiotic fiber are individualized
Work from the labs of Dr. Justin Sonnenberg and Dr. Purna Kashyap published in 2016 sought to characterize the individual variation in response to a dietary intervention.24 Using three healthy human donors, they colonized germ-free mice with human fecal material and allowed the microbial community to equilibrate before manipulating the amount of prebiotic fructo-oligosaccharide (FOS). Interestingly, the magnitude of change in the fecal metabolome following the introduction of FOS to the humanized mice was unrelated to the magnitude of change in the composition of the gut microbiota. The authors concluded that compositional changes affected by diet do not predict community functionality, and that assessing an individuals’ changes in microbiota function will be necessary for precision microbiota-directed dietary interventions.
At least one human study has attempted to do this with targeted metabolomics. Venkataraman et al. (2016) found variable fecal butyrate responses in healthy young adults to dietary supplementation with resistant starch.25 In this case, the microbiota of individuals that responded to resistant starch had an elevated abundance of the butyrate producing microbe Eubacterium rectale.
Baseline microbiota can predict blood glucose response to foods
Other groups have utilized baseline gut microbiota characteristics, combined with host parameters, to predict acute dietary responses. For instance, Zeevi et al. (2015) combined metagenomic sequencing with machine learning to predict blood glucose responses to individual foods and mixed meals with greater accuracy than glycemic index or carbohydrate counting.26 Moreover, when they put diabetic or pre-diabetic patients on a personalized nutrition plan based on the predictions for three months, 93 percent had a reduction HbA1c, a marker of long-term blood glucose control, and forty percent had a clinically significant decrease in HbA1c. In the next few years, we’ll likely see a lot more integration of machine learning with microbiome data to inform optimal dietary choices.
The future of diet-microbiome modulation is personalized
A study published by Dr. Dan Knights lab at UC Denver in 2019 performed detailed analysis of the fecal metagenomes and diet records of 34 healthy human subjects collected over 17 days.27 They found that gut microbiota composition depended on at least the prior two days’ dietary intake. Overall, diet accounted for an estimated 44 percent of the total variation in average microbiome composition. While the majority of subjects had significant food-microbe interactions, very few of these were conserved across individuals. Again, these data suggest the need for considering individuality as we design dietary strategies to modulate the microbiome to improve health!
Conclusion & future directions
Hopefully you enjoyed this summary as much as I enjoyed writing it! These papers have certainly shaped the way I think about diet and nutrition, and how we might use diet to beneficially modulate our microbial communities in states of health and disease. I’ll be sure to update this article as new research comes out and as we continue to elucidate these complex host-microbe-nutrition interrelationships.
In part 2 (stay tuned!), I will dive deeper into the hologenome concept, challenge some basic assumptions, and discuss some of today’s most popular dietary practices and how they impact the gut microbiome.
That’s all for now! If you appreciated this article, be sure to leave a comment below, subscribe to my newsletter, or consider supporting the blog directly by becoming a patron.