Targeting the Gut Microbiota’s Role in Host Energy Absorption With Precision Nutrition Interventions for the Prevention and Treatment of Obesity

Abstract

The field of precision nutrition aims to develop dietary approaches based on individual biological factors such as genomics or the gut microbiota. The gut microbiota, which is the highly individualized and complex community of microbes residing in the colon, is a key contributor to human physiology. Although gut microbes play multiple roles in the metabolism of nutrients, their role in modulating the absorption of dietary energy from foods that escape digestion in the small intestine has the potential to variably affect energy balance and, thus, body weight. The fate of this energy, and its subsequent impact on body weight, is well described in rodents and is emerging in humans. This narrative review is focused on recent clinical evidence of the role of the gut microbiota in human energy balance, specifically its impact on energy available to the human host. Despite recent progress, remaining gaps in knowledge present opportunities for developing and implementing strategies to understand causal microbial mechanisms related to energy balance. We propose that implementing rigorous microbiota-focused measurements in the context of innovative clinical trial designs will elucidate integrated diet-host-gut microbiota mechanisms. These mechanisms are primed to be targets for precision nutrition interventions to optimize energy balance to achieve desired weight outcomes. Given the magnitude and impact of the obesity epidemic, implementing these interventions within comprehensive weight management paradigms has the potential to be of public health significance.

INTRODUCTION

Obesity is a disease that affects populations around the globe, with the United States having 1 of the greatest burdens of disease. Recent data-driven models estimate the prevalence of obesity in the United States through 2021 to have been 172 million adults aged older than 25 years. Forecast models indicate that by 2050, 213 million adults will have overweight and obesity.¹ Because of the magnitude of disease and the associated downstream negative health outcomes, obesity requires a multipronged strategy for prevention and long-term management.² The primary physiological driver of body weight regulation is energy balance,³ which refers to the relationship between energy intake (ie, available energy from consumed diet) and energy expenditure. Positive energy balance stemming from an energy intake that exceeds energy expenditure leads to weight gain if sustained over time.⁴

There are many biological modulators of energy balance, 1 of which is the gut microbiota.⁵ The gut microbiota is defined as the collection of living microorganisms primarily residing in the colon, whose genetic components and the resultant functions (ie, gut microbiome) play important roles in human physiology.^6,7 The gut microbiota has emerged as a direct contributor to energy balance in mice primarily via its key role in harvesting energy from the host dietary components that reach the colon undigested.^8,9 Evidence to date points to this being the primary mechanism by which the gut microbiota modulates energy balance because the energy harvesting capacity of the colonic microbes directly affects the energy that the human host ultimately absorbs.^9–11 However, evidence also suggests a potential role of the gut microbiota in energy expenditure^12,13 and energy intake.^11,14

Translation of this intriguing body of preclinical evidence to humans has been largely restricted to cataloguing differences in gut microbiota composition, comparing people with obesity to those without obesity or in response to weight loss interventions.^15–17 Most clinical studies have lacked the appropriate controls or precise and comprehensive assessments of the entire energy balance equation. This has greatly limited the uncovering of causal mechanisms that attribute changes in energy balance to specific functions of the gut microbiota.

Overview of Human and Microbial Bioenergetics: Relevance to Energy Balance

The fate of energy harvested by the gut microbiota from host dietary components that escape digestion and absorption in the small intestine is the key bioenergetic principle linking diet, microbes, and the energy ultimately available to the human host. This microbially derived energy is produced primarily via fermentation of carbohydrates from host diet and, to a lesser degree, proteins. About 90% of calories from the host diet are absorbed in the small intestine, with values ranging from ∼83% to 97% (Figure 1).^14,18–23 When undigested dietary components reach the colon and undergo microbial fermentation, most of the resultant energy is either absorbed by the host, used to grow microbial biomass, or excreted in feces. Energy absorption by the host is commonly approximated as a percentage of consumed energy after accounting for loss of fecal energy that is incompletely used by the host and microbes, and this is called metabolizable energy.²⁴ Relevant to precision nutrition, fecal energy loss varies across individuals (2%-16%) due to methodological and biological factors.^14,18–23

A fundamental gap in knowledge is a precise quantification of how much of the energy produced by colonic microbes ultimately contributes to host energy stores. Expanding our understanding of these bioenergetic processes represents an important opportunity for developing individualized approaches to alter energy balance via the gut microbiota.

Prevailing assumptions about dietary energy absorption in the small and large intestine are based on established nutritional constants that do not account for the variable role of the colonic microbiota in dietary energy harvest.²⁵ Importantly, the mechanisms that control the proportion of energy absorbed from the colon vs excreted in feces as undigested food, microbial biomass, or byproducts of fermentation are not precisely defined. This tug of war or collaboration between host and microbes for energy and nutrients could be important for weight management because of its potential impact on energy balance.¹⁸ Thus, the fate of colonic energy and the mechanisms determining its flux is an important gap that must be addressed by precisely quantifying host and microbial bioenergetics in vivo.

Critical Interactions Among Host, Diet, and the Gut Microbiota: An Opportunity for Precision Nutrition

Given the strong rationale for a role of the gut microbiota in modulating the dietary energy available to the human host (ie, metabolizable energy), obesity stands out as a viable target for microbiota-directed precision nutrition. To achieve this, a deep mechanistic understanding of the complex interactions among the human host, diet, and the colonic microbiota is essential for developing microbiota-focused therapeutics targeting energy balance.²⁶ Three fundamental interactions are relevant. First, colonic microbes depend on host diet for energetic substrates; thus, precision diets aimed to optimize both host health and microbial metabolism are needed.²⁷ Second, various facets of host energy balance are affected by microbial metabolites produced through fermentation of these dietary substrates and host intestinal mucus.²⁶ These facets include energy absorption, energy expenditure, and energy intake.²⁸ Third, host biological factors, such as genetics and metabolites, influence the interindividual variation in gut microbial functions and need to be considered via precision nutrition.²⁹

Because diet is arguably the most fundamental modulator of the gut microbiota,^30,31 increasing the proportion of dietary components delivered to the colon is a key strategy for remodeling the composition and function of the gut microbiota. Diets consumed by the host can range from being very low to very high in substrates for colonic microbial fermentation. A standard Western diet essentially “starves” the gut microbiota of its preferred substrates.³² On the other hand, a high-fiber, whole-foods diet with lower digestibility in the small intestine provides an abundance of substrates for gut microbial growth and activity (primarily fermentation).³² Precision nutrition approaches that account for an individual’s microbial characteristics and host physiology can exploit these fundamental interactions between dietary components and the gut microbiota to influence host energy balance. Given that energy absorption is understudied in humans, yet is likely to directly affect energy balance,¹⁸ we review here this component of energy balance, focusing on recent clinical findings that have advanced knowledge about host-diet-gut microbiota interactions the affect human bioenergetics.

METHODS

To identify literature for this review, we considered clinical trials of dietary interventions whose results were published within the past 5 years. The included articles reported on trials that evaluated energy balance related to energy absorption in the context of obesity, with consideration of the gut microbiota. Publications were identified in PubMed using the terms (((gut microbiome) AND ((energy balance) OR (obesity) OR (energy absorption)) with the filters clinical trial, randomized controlled trial, English Language, humans, and published in the past 5 years (through March 19, 2024).

The resultant abstracts (N = 192) from the identified studies were reviewed for relevance and selected for discussion (n = 4). Criteria for exclusion included the lack of a diet intervention (supplement, probiotic, single food, single ingredient, and studies involving pharmacological agents were excluded) or the lack of evaluation of metabolizable energy, energy absorption, or fecal energy loss. Although the focus of this review is on the role of the gut microbiota in energy availability to the host, clinical trials that measured energy intake or energy expenditure were considered for inclusion if they provided valuable supporting evidence. To ensure that no critical articles were missed due to the use of the term “gut microbiome” instead of “gut microbiota,” an additional search that included both terms was performed using the same terms approach as the original search. That search yielded 225 total abstracts (n = 33 additional abstracts). No other abstracts were identified as suitable for inclusion.

Three additional searches were conducted with the same filters to ensure key literature was not missed in any energy balance phenotype. For fecal energy loss or energy absorption, the following search terms were used: ((“gut microbiome”) AND (“energy absorption”)) OR (“fecal energy loss”), which yielded 1 abstract. For energy expenditure, the following search terms were used: (“gut microbiome”) AND (“energy expenditure”), which yielded 4 abstracts. For energy intake, the following search terms were used: (“gut microbiome”) AND (“energy intake”), which yielded 10 abstracts.

One final search was conducted to capture clinical trials evaluating the microbiome, energy balance, and energy absorption or fecal energy loss without a filter for time, meaning that all available literature was searched. The search terms used were (((“gut microbiome”) AND (“energy balance”)) OR (“energy absorption”)) OR (“fecal energy loss”), with the following filters: clinical trial, randomized controlled trial, English, humans. This yielded 53 abstracts. The additional searches identified 1 additional article to include in this review.

Given that the evidence in this space is limited, the “Conclusion” section is used to build a framework for advancing the field. The literature cited is based on the collective expertise of the authors.

DISCUSSION

Context and Rationale

Several lines of evidence, mainly in mice, suggest that the channeling of energy away from the host, primarily via fecal energy losses, could contribute to variability in body weight regulation.¹⁸ Given the role of the gut microbiota in modulating energy available to the human host,⁹ it is plausible that this capacity can be exploited to modulate body weight. Early rationale for this possibility in humans was revealed with 2 approaches: overfeeding and/or underfeeding³³ and administration of antibiotics to isolate the role of the gut microbiome in energy extraction.³⁴

First, with an overfeeding paradigm, Jumpertz et al³³ provided early clinical evidence of the relationships among diet, the gut microbiota, and energy absorption. Men without (n = 12) and with (n = 9) obesity were provided 2 fixed energy diets (2400 vs 3400 kcal day^–1) in random order. The interventional periods were flanked by weight-maintaining diets.³³ The fixed energy content meant that the degree of overfeeding relative to energy requirements varied between people. Among all participants, overfeeding (3400 kcal day^–1) led to rapid changes in the relative abundance of several gut bacterial taxa as compared with the weight-maintaining diet. Overfeeding was associated with less energy lost in feces as compared with underfeeding, but only in individuals without obesity, despite similar gut microbiota composition in individuals with and without obesity.³³ That fecal energy losses were different in people with obesity as compared with those without, in light of rapid changes to the gut microbiome in response to overfeeding, aligns with evidence in mouse studies showing that the colonic microbiota affects energy availability to the host.^9,11

Next, Reijnders et al³⁴ used antibiotics in an early attempt to translate the preclinical evidence linking the absence of a gut microbiota (germ-free mice) with lower fat-mass gain. Germ-free mice are often used to isolate the specific role of the gut microbiota because they are bred in a sterile environment free of bacteria and other microorganisms.³⁵ Using antibiotics in humans creates a similar model for isolating the role of the gut microbiota in energy balance. The study authors administered a 7-day antibiotic treatment (vancomycin or amoxicillin) to 57 men with obesity.³⁴ Compared with placebo, vancomycin, but not amoxicillin, led to a distinct remodeling of the gut microbiota and reductions in both fecal short-chain fatty acids (SCFAs) and bile acids.³⁴ However, several key metabolic end points, including fecal energy, did not change. A major confounder of these results was the lack of a controlled diet, because variability in diet digestibility in both conditions may have obscured differences in fecal energy loss. These 2 early studies set the stage for the recent trials that have yielded important new insights into host-diet-microbiome interactions influencing energy balance.

Recent Clinical Studies

One gap remaining in the early literature was the lack of customization of caloric intake to the energy needs of the host. To address this, Basolo et al³⁶ implemented 3 days of proportional over- or underfeeding (50% or 150% of weight-maintaining calories, respectively) in men and women with obesity. Diets were equivalent in nutrient composition and were provided in a randomized crossover manner with a 3-day washout period.³⁶ This initial study phase was followed by a randomized, double-blind, parallel-arm study of vancomycin vs placebo with a weight-maintaining diet. Daily fecal energy was evaluated via a combination of bomb calorimetry and fecal marker dyes. The calories provided were calculated based on equations developed in the authors’ clinical research unit. They found that during underfeeding, fecal energy loss as a percentage of caloric intake (metabolizable energy) was higher, whereas the absolute fecal energy loss was lower due to a smaller energy load, compared with the overfeeding condition.³⁶ Although the gut microbiota was relatively stable within individuals, the relative abundances of Akkermansia muciniphila, Bacteroides coprocola, Lachnospiraceae, and Ruminoccoccus spp. were different between the over- and underfeeding conditions independent of diet order. This suggests a relationship between change in these taxa and differential energy absorption.

In the second phase of the study, participants randomized to receive vancomycin treatment had higher fecal calorie losses (83.3 more kcal day^–1) compared with placebo, meaning the gut microbiota may be necessary for making energy available to the human host, as previously shown in mice.³⁶ There was no difference in energy expenditure or substrate oxidation between study arms. Although analyses lacked a comparison to people without obesity and did not explore the potential effects of differences in diet composition, this study provided evidence of a correlation between the gut microbiota and human energy absorption.

An advancement in the use of equations to estimate energy expenditure was to match energy intake to measured energy expenditure.³³ Yoshimura et al³⁷ addressed this by conducting a randomized crossover feeding study over 8 days (4 days, 3 nights for each outpatient and inpatient; >3-week washout between diets) in which participants were fed based on their energy expenditure, measured by triaxial accelerometry.³⁷ Young men who were lean and otherwise healthy underwent a control condition in which they were fed at energy balance, an underfeeding condition (caloric intake ∼75% of energy expenditure), and an overfeeding condition (150% of energy expenditure). The 3 conditions included the same food items and were macronutrient-matched (15% protein, 30% fat, and 55% carbohydrates), with energy content determined based on food labels. Gut transit was assessed using dye markers. Food and fecal energy were assessed via bomb calorimetry. Daily energy losses in feces, urine, and the combination of feces and urine were calculated as a percentage of energy intake and termed “energy excretion rate.”³⁷ The study found no difference in fecal energy excretion rate across diet conditions. However, overfeeding was associated with alterations in α-diversity and gut microbiota composition at the phylum and genus levels as compared with the control, even in this short 8-day period.³⁷ Although it is surprising that rate of fecal energy loss was not altered by overfeeding, the short feeding duration may have been insufficient, or a lack of precision in the fecal energy measurement may have reduced the signal to noise ratio.

Although Jumpertz et al³³ showed that calorie-restricted diets alter gut microbial community structure and fecal energy, the magnitude of change can vary based on the selected dietary intervention. In addition, whether the total amount and type of dietary energy (ie, macronutrients, including fiber) available to gut microbes affects energy balance is still an open question. To address these gaps, von Schwartzenberg et al³⁸ used a liquid meal replacement diet which is an effective means to reduce body weight³⁹ but is low in nutrients accessible to the microbiota for fermentation (ie, dietary fiber).³⁸ Eighty postmenopausal women with overweight or obesity first followed an 8-week, very-low-calorie liquid diet (VLCD; Optifast II, Nestlé, Vevey, Switzerland; 800 kcal day^–1; 10.8 g day^–1 fiber, typically from inulin), followed by 4 weeks of a conventional reduced-calorie diet (1616 ± 648 kcal and 27.3 ± 15 g day^–1 dietary fiber; mean ± s.d.), and, finally, 4 weeks of a weight-maintaining diet (1833 ± 569 kcal and 26.9 ± 14 g day^–1 dietary fiber; mean ± s.d.).³⁸

The combined VLCD and reduced energy intervention led to significant weight loss (13.6% ± 4.0% over 12 weeks; mean ± s.d.). Glucose regulation improved during the VLCD only. However, the VLCD, as compared with a higher[fiber, reduced-calorie diet, led to a reduction in 16S RNA gene copies,³⁸ which approximate microbial biomass that could alter energy balance.¹⁴ In addition, there was a marked disruption of gut microbial structure and inferred metabolic activity. Fermentation capacity was decreased, as evidenced by reduced branched-chain amino acids and SCFAs, despite an increase in the relative abundance of microbial taxa generally considered responsible for generating these metabolic byproducts.³⁸ This was likely because of the reduction in microbial biomass, arguing for the critical importance of absolute quantification of microbial species.^40–42 Although fecal energy was not measured in these participants, germ-free mice that were inoculated with human feces collected after the reduced-calorie intervention had lower fecal energy (assessed by bomb calorimetry; kcal g^–1 dry fecal mass) as compared with those inoculated with pre-diet feces. Furthermore, post-diet fecal microbiota transplantation (FMT) transferred the phenotype of reduced adiposity to recipient mice relative to pre-diet FMT, suggesting that the gut microbiota was at least partially mediating the diet-induced energy balance alterations.³⁸

Our group recently addressed a fundamental question: What is the quantitative contribution of the gut microbiome to human energy balance?¹⁴ In contrast to prior studies, we used a quantitative bioenergetics paradigm to precisely measure the entire energy balance equation (energy intake, energy expenditure, and fecal energy loss) in the context of eucaloric and isocaloric high-fiber, whole-foods diet as compared with a low-fiber diet. Dietary energy needs were determined via energy expenditure measured with whole-room calorimetry, and all foods were prepared in a metabolic kitchen using research-grade software to evaluate nutritional composition.⁴³ Under tightly controlled dietary and environmental conditions, we implemented a randomized crossover, controlled feeding study in which participants were fed over 11 days at home and 11 days while domiciled in our metabolic ward with a > 14-day washout in between diet conditions. The primary outcome was metabolizable energy,²⁴ and both host and microbes were deeply phenotyped.⁴³

We found that the overall efficiency of energy absorption (ie, metabolizable energy) was substantively reduced on a high-fiber, whole-foods diet relative to a low-fiber diet, which was coupled to robust increases in gut microbial biomass.¹⁴ We observed substantial between-person variation in metabolizable energy only when microbes had ample access to fermentable, diet-derived substrates, supporting the hypothesis that one’s unique gut microbial ecosystem modulates fecal energy loss. Energy expenditure was not different between diets. Using metagenomic sequencing, we found that both β-diversity and the composition of the gut microbial ecosystem were dramatically different between diets.¹⁴ Critical to our understanding of mechanisms driving energy balance, we uncovered 3 factors that explained the interindividual variability in metabolizable energy that occurred with the high-fiber, whole-foods diet only (range, 84%-96% on the high-fiber diet): 1) microbial biomass, 2) microbial fermentation, and 3) colonic transit time (Figure 2). We posit that this variation could drive individualized propensity for weight gain.

Our findings suggest a potential unifying paradigm because they align with the original concept, proposed by Neel, ⁴⁴ that a thrifty genotype is a biological driver of the propensity for obesity. This was conceptually extended to the gut microbiota by Fraser-Liggett and Shuldiner,⁴⁵ who posited that a thrifty gut microbial community may be a disadvantage in the context of excess food availability by promoting energy absorption. Our findings are the first, to our knowledge, to provide evidence that the gut microbiota is 1 component that modifies metabolic thriftiness through its impact on energy absorption from diet. We posit that less metabolic thriftiness leading to reduced energy absorption is only revealed when gut microbes have ample access to their preferred substrates from the host diet. This raises the possibility that individuals with a thriftier metabolic phenotype, of which a thrifty gut microbiota is 1 component, have a propensity for positive energy balance and weight gain, even when consuming a healthful diet (Figure 3).¹⁴ Overall, this study raises the hypothesis that energy balance can be modulated via host-diet-gut microbiome interactions that alter energy absorption from consumed diet. These concepts should be considered in future clinical trials and are primed for modulation via precision nutrition.

It is plausible that specific dietary patterns could alter the energy-harvesting capacity of the gut microbiota in addition to or independently of over- or underfeeding. One example is time-restricted eating (TRE), which is a type of intermittent fasting that shortens the eating window. The shortened eating window can be aligned with the circadian clock or implemented anytime during the day.⁴⁶ Various studies have shown that TRE affects body weight,^47,48 but mechanisms are unclear. Dawson et al²³ tested the effects of early TRE (eTRE) on energy absorption in a randomized crossover, controlled feeding study enrolling young, generally healthy individuals (N = 16 who completed the study). Participants were fed a weight-maintaining diet for 9 days that was prepared in a metabolic kitchen and contained 60% carbohydrate, 16% fat, 28% protein, and 37.7 g of fiber. Calorie levels were based on resting energy expenditure measured via metabolic cart and a physical activity factor based on self-reported activity levels. The diets were consumed outpatient, with supervised consumption of breakfast on 4 of the study days. The eTRE window was between 8 am and 2 pm, thus aligning with circadian rhythms. The control eating window was from 8 am until 8 pm. Based on 3 days of fecal collection (using dye markers to estimate the fecal collection period), Dawson et al found no difference in fecal energy loss between eTRE and the control. Nonetheless, there was interindividual variability in energy absorption, potentially due to the high measurement error of dye markers and/or the potential for suboptimal diet adherence in the outpatient setting. Early TRE altered the composition of the gut microbiota and improved cardiometabolic markers relative to the control diet.²³ The impact of TRE on energy balance via gut microbiome alterations is plausible, yet multiple gaps remain. The specific time window needed to maximize benefit, a precise investigation of the entire energy balance equation, and the role of dietary composition and caloric restriction require direct investigation. Such interventions would ideally be conducted in a metabolic ward setting to optimize adherence, thus ensuring delivery of intended nutrients to the gut microbiota. Additional considerations for future studies are to evaluate not only gut microbiota composition but also microbial metabolites, investigate the functional mechanisms by which the gut microbiota might mediate the effects of diet on energy balance, and maximize the capacity of the microbiota to modulate energy balance by comparing shortened eating windows that are low vs high in microbiota-accessible carbohydrates. Importantly, other dietary patterns, such as ketogenic⁴⁹ or Mediterranean diets,⁵⁰ could yield beneficial energy balance effects, and they warrant future study.

Potential Mechanisms Driving the Gut Microbiota’s Role in Human Energy Balance

Taken together, the studies discussed herein suggest that microbial determinants of metabolizable energy may predispose some individuals to weight gain. In the reviewed studies, these microbial determinants were modulated by underfeeding or overfeeding the host,^33,36 reducing the totality of the host’s gut microbial biomass with antibiotics,^34,36 providing diets with foods that are accessible to the gut microbes,^14,38 or via altering the timing of eating with intermittent fasting approaches.²³ Despite this recent progress, several open questions remain that lead to intriguing testable mechanistic hypotheses (Figure 4).

Open Question 1: What Is the Quantitative Flux of Energy Between the Human Host and Colonic Microbes?

The absorption of dietary energy in the small intestine determines the proportion that can reach the colon. The energy that reaches the colon is then available to the gut microbiota for fermentation. The ultimate fate of the energy harvested by gut microbes is not fully understood, and this fate is a key component of energy balance (Figure 1). Data from mouse studies show that the gut microbiota is necessary for maximizing host energy absorption, as illustrated by comparing energy balance components in germ-free mice and conventionally raised mice that harbor a microbiota. Germ-free mice have smaller fat pads than germ-free mice that are subsequently inoculated with the cecal contents of conventionally raised mice, despite greater food intake and lower energy expenditure. Therefore, the role of the gut microbiota on adipose stores was primarily driven by higher fecal energy losses in germ-free mice compared with conventionalized mice. The higher energy absorption in conventionalized mice was accompanied by the triggering of de novo lipogenesis upon monosaccharide absorption and by adipose tissue expansion due to suppression of an inhibitor of lipoprotein lipase (fasting-induced adipose factor).¹¹ In addition, mouse models of obesity lose less fecal energy than those without obesity, and this may be due to compositional differences in the gut microbiota. When germ-free mice are colonized with microbes from lean vs obese donors, the donor’s weight phenotype is recapitulated in recipient mice.⁹ This suggests that 1 of the mechanisms that promotes weight gain in rodent models is increased absorption of the energy harvested by the gut microbiota. This opens up the possibility of a gut microbiota-adipose tissue axis that can be studied in humans via adipose tissue biopsy specimens to test both endocrine and exocrine mechanisms.

An intriguing set of hypotheses has emerged in humans related to the quantitative flux of energy between humans and microbes. We demonstrated that the negative energy balance achieved by remodeling the gut microbiota with a high-fiber, whole-foods diet is driven, at least in part, by the shunting of energy from microbial activity in the colon toward the growth of microbial biomass instead of toward host energy stores.¹⁴ This is likely a quantitative driver of energy balance, because our mathematical model¹⁴ and prior literature^51,52 suggest that 25%-50% of fecal energy is derived from microbial biomass. Indeed, the relationship we uncovered between microbial biomass and variation in energy absorption on a high-fiber, whole-foods diet only¹⁴ puts forth biomass growth as a mechanism to alter energy balance via precision nutrition. To uncover targetable mechanisms, studies are needed to quantify the bidirectional flux of energy between microbes and host and, importantly, establish the biological physiological, and sociodemographic factors that affect interindividual variability.

Open Question 2: Do Host Physiological Factors Drive Variability in Energy Absorption?

A fundamental physiological factor that could drive variability energy absorption is colonic transit time. Colonic transit time is highly variable between individuals and affects gut microbial metabolism⁵³ because it controls the length of time colonic microbes have to ferment dietary substrates. In mice that have been depleted of gut microbes, which have drastically reduced production of SCFAs, there are higher levels of GLP-1.⁵⁴ This leads to slowing of transit through the small intestine, which, in turn, generally slows colonic transit,⁵³ and higher energy absorption. Thus, in conditions where microbes are absent, colonocytes sense the reduced energy availability from SCFAs and regulate GLP-1 secretion to accommodate this energetic state.⁵⁴

Interactions between the gut microbiota and colonic transit may also be important for human energy absorption. We found that although colonic transit time was not directly related to metabolizable energy, it was a critical factor in capturing between-person variation in metabolizable energy, according to our mathematical model.^14,55 Studies have also shown that colonic transit time is associated with fecal energy losses, whereby lower fecal energy losses occurred in people with faster transit times.⁵⁶ In addition, in a study comparing diets high vs low in microbiota-accessible carbohydrates, colonic transit time was not different by diet but did explain 5% of gut microbiome variation.⁵⁷ Because the function of the gut microbiota drives energy harvest, it is plausible that colonic transit time is a physiological factor that may affect energy absorption by the host.

Other gut physiological characteristics could impact energy absorption. Gastric emptying controls the rate of delivery of nutrients to the small and large intestine, and as such, affects satiety and body weight.⁵⁸ Thus, it could conceivably affect energy absorption of microbially derived energetic substrates. However, a role of gastric emptying in energy absorption has not been identified in rodents⁵⁴ or humans.¹⁴

A thicker colonic mucus layer could lead to lower energy absorption by the host.⁵⁹ This aligns with the findings that Western diets are associated with higher energy absorption and a microbial community that is more likely to use the mucus layer as an energy source.^14,32 Given that a depleted mucus layer is associated with higher intestinal permeability,⁶⁰ this may be the mechanism that facilitates a greater absorption of energetic substrates.⁶¹

An intriguing set of findings in mice demonstrated that cold exposure, with subsequent transplantation of the “cold microbiota” to germ-free mice, led to increased adipose beiging or thermogenesis, improved insulin sensitivity, and an increase in intestinal surface area. These alterations, particularly increased intestinal surface area, may increase host energy absorption to meet the higher energy demands during cold exposure.⁶² Given the wide interindividual variability in host energy absorption, studies should be conducted to investigate whether these gut-related mechanisms are mediators of energy balance that could be targeted by precision nutrition.

Open Question 3: Do Certain Microbial Species Alter Energy Absorption in the Human Host?

Several bacterial taxa have been identified to have a relationship to energy absorption^10,14,36,37 or obesity.⁶³ The mechanisms driving these associations are unclear in humans because they rely on relative abundance measures of microbial composition and lack an exploration of community-wide functional interactions. Thus, there is currently no consensus on the gut microbiota composition that will optimize energy absorption. Importantly, an “optimal” gut microbiota is likely to be context-dependent and thus vary substantially between individuals.

There are many intriguing mechanisms by which specific microbes in model systems affect energy absorption, but they have yet to be translated to humans. These preclinical mechanisms have been reviewed extensively recently,^64,65 and a few salient examples are discussed herein. In ob/ob or diet-induced obesity models, treatment with Parabacteroides distasonis reduced weight gain via reduced food intake. This is considered a core microbial species whose relative abundance is lower in people with obesity.⁶⁶  Akkermansia muciniphila is a well-studied species that has been associated inversely with obesity in multiple model systems.⁶⁷ Treatment of obese mouse models with A. muciniphila reverses high-fat-diet-induced increases in weight and fat mass, potentially via reduction of inflammation and restoring gut barrier integrity.⁶⁸ A unique aspect of A. muciniphilla’s mechanism of action is that a surface protein (postbiotic) confers many of its metabolic actions.⁶⁹ The energy balance mechanisms of action by which A. muciniphila protects against obesity include increased energy expenditure and fecal energy loss.⁷⁰ In addition, the relative abundance of Bacteroides thetaiotaomicron is lower in people with obesity. Treatment of mice with live, but not heat-killed, B. thetaiotaomicron lowers fat mass and increases lean mass.⁷¹  Methanobrevibacter smithii, the most common methanogen in the gut, seems to be associated with energy extraction,^72–74 apparently through interaction with bacteria that produce SCFAs.⁷⁴

Recently, our group developed a first-in-human method to measure methane over 24 hours.⁷⁵ We found that an very small amount of methane was expelled as calories. Therefore, although methanogenesis does not appear to be a quantitative driver of energy balance, abundance of M. smithii and the associated methane expelled may be biomarkers of fermentation that correlate with energy balance. A critical way forward to translating these findings into precision therapeutics requires understanding community-wide functional interactions, because microbial niches are complex ecosystems in which defined competition and synergy coexist.

Open Question 4: Do Microbial Metabolites Alter Energy Absorption in the Human Host?

Fecal and circulating metabolites can serve as evidence of host-diet-microbiome interactions that affect energy balance.⁶⁴ It is unlikely that microbially derived metabolites substantively affect energy absorption directly, because the energetic content lost in feces is expected to be small.⁵⁵ However, there could be indirect roles in energy balance related to signaling functions or because these metabolites represent biomarkers of fermentation.

For instance, the signaling functions of SCFAs are particularly well established in preclinical models⁷⁶ and are believed to affect adipose stores via increased expression and signaling from G-protein coupled receptors 41 and 43.⁷⁷ Short-chain fatty acids may also trigger a switch from lipogenesis to fat oxidation via a PPAR-γ driven signals.⁷⁷ However, there is controversy about whether individuals with obesity produce more or fewer SCFAs, making the mechanisms in humans unclear.⁷⁸ This could be because most studies have lacked a controlled diet, which provides the inputs for SCFA production; instead, they have relied on measures at a single time point instead of flux or 24-hour excretion in feces, and potentially interindividual variability in the biological effects of SCFAs. Using a controlled diet, we showed that fermentation, as exemplified by the 24-hour concentration of the fecal SCFA propionate, was associated with human metabolizable energy interindividual variation when providing a diet high in microbiota-accessible carbohydrates.¹⁴ The most direct evidence of a role of SCFAs in human energy balance came from a study from Canfora et al.⁷⁹ They administered to men with obesity (n = 12) colonic infusions of SCFA mixtures at concentrations and ratios similar to what would be achieved with a high-fiber diet. They uncovered increases in fat oxidation, energy expenditure, and PYY, along with increased lipolysis as a result of SCFA colonic infusion.⁷⁹ Although fermentation products like SCFAs are energy sources that are readily absorbed by other microbes and the host,⁷⁶ the quantitative impact on energy balance is unknown and likely to vary across individuals.

A particularly well-studied class of metabolites is primary and secondary bile acids, the latter of which are produced by gut microbial metabolism. In addition to the role of bile acids in absorption of dietary fats, they also signal through farnesoid X receptor and Takeda G protein-coupled receptors 5 to promote negative energy balance.^80,81 The role of P. distasonis on prevention of obesity, as discussed under question 3, is mediated via production of secondary bile acids and succinate.⁶⁶ Bile acid mechanisms that alter energy balance are likely active in humans because there are correlations between altered bile acid profiles and obesity.⁸² Given the interaction between host and microbes in modulating the bile acid pool, it is not surprising that causal inferences have been difficult to establish.

There are several other mechanisms by which host-microbiota co-metabolites could affect energy balance based on in vitro, cell culture, or animal models. For example, the relationship between B. thetaiotaomicron and higher body mass index in humans is also inversely correlated with higher levels of glutamate. B. thetaiotaomicron, a human commensal, ferments glutamate, so this association suggests a potential mechanistic pathway.⁷¹ Plant-based metabolites like trigonelline (N-methylnicotinate) serve as catabolic substrates for gut microbes and, therefore, may fuel microbial biomass growth.⁸³ Microbial biomass growth, in turn, could contribute to fecal energy loss. Trigonelline also promotes adipocyte browning in mice by increasing the expression of β3 adrenergic receptors in white adipose tissue, potentially contributing to thermogenesis.⁸⁴ Other metabolites reflecting a scarcity of microbiota-preferred energetic substrates may promote a positive energy balance. These include byproducts of protein catabolism such as p-cresol and simple sugar catabolites such as fucose that may limit biomass growth⁸⁵ and promote ectopic fat redistribution.⁸⁶ Taken together, studies are needed to identify more bona fide microbial metabolites within metabolomics platforms, quantify how the flux of these metabolites could affect energy balance directly or indirectly, uncover causal mechanisms, and, ultimately, translate this knowledge to human paradigms.

Open Question 5: What Is the Role of the Gut Microbiota in Human Energy Expenditure?

The gut microbiota could additionally affect body energy stores via changes in expenditure. One mouse study found that cecum removal promoted weight gain relative to a sham procedure due to reduced energy expenditure by the host (indirect calorimetry) and gut microbes (direct calorimetry).¹³ In this experimental paradigm, dietary or pharmacological manipulation of the gut microbiota altered anaerobic energy expenditure. Importantly, all the differences noted in energy expenditure were not detected with indirect calorimetry and, thus, were due to anaerobic thermogenesis.¹³ This suggests a role of gut microbiota composition in modulating energy expenditure. To date, human studies have not found a relationship between the gut microbiota and energy expenditure.^14,36 This could be because the magnitude of anaerobic microbial thermogenic processes in humans is not known. Microbial thermogenesis cannot be measured with the indirect calorimetry methods currently used to measure human energy expenditure.¹³ Novel methods must be developed to determine whether the gut microbiota contributes to thermogenesis in humans.

Open Question 6: What Is the Role of the Gut Microbiota in Ingestive Behaviors?

Ingestive behaviors also are modulated by the gut microbiota, with the most robust data being from animal models. Microbial metabolites of dietary macronutrient degradation (ie, SCFAs and bile acids) stimulate the release of satiety hormones such as GLP-1 and PYY from intestinal cells and through vagal signaling mechanisms.⁸⁷ In addition, the gut microbiota can produce neurotransmitters such as serotonin and dopamine.⁸⁸ Because of the robust, bidirectional communication between the gut and brain, it is highly likely that, in certain contexts, the gut microbiota may affect energy balance by altering satiety and food intake.⁸⁷

In humans, there is consistently replicated observational data showing that diets high in prebiotics or prebiotic supplements improve satiety and energy intake.⁸⁷ We demonstrated that remodeling the microbiome under energy balance conditions alters the levels of several hormones, including GLP-1, leptin, and pancreatic polypeptide, which are related to ingestive behaviors, without an impact on satiety or food intake.¹⁴ Whether these satiety signals translate into alterations in ingestive behaviors in other conditions, such as during caloric restriction, remains to be tested. The collective body of clinical literature has yet to elucidate causal mechanisms or the quantitative impact on energy balance that is directly related to the gut microbiota.

CONCLUSIONS

Gaps and Opportunities for Precision Nutrition Interventions Designed to Remodel the Gut Microbiota to Optimize Energy Balance

Translating the vast body of preclinical and clinical evidence regarding the role of the gut microbiota in obesity to actionable clinical and public health recommendations is hindered by the limited evidence in humans. There are several fundamental open questions: 1) Is the gut microbiota on the causal pathway between diet and human obesity? 2) If the gut microbiota contributes causally to obesity, what is the magnitude and variability of the effect size? 3) How do obesity and other health conditions (eg, diabetes and metabolic dysfunction–associated steatotic liver disease) affect energy absorption? 4) Do caloric restriction or overfeeding alter energy absorption, and if so, what are the mechanisms and how do they differ in individuals with and without obesity? And 5) which specific dietary components, beyond microbiota accessible carbohydrates, affect energy balance via the gut microbiota? Several high-priority research strategies that are poised to address these important gaps are discussed next.

Quantitative Microbiome Measurements

An important hindrance in quantifying effect sizes relating the gut microbiota to obesity is that most studies rely on relative abundance measurements to establish correlations between gut microbial composition and human phenotypes. Although such methods are useful for determining the proportional abundance of major taxa within the colonic microbiota, they do not always allow for interpreting the direction or magnitude of change in taxonomy over the course of an intervention.⁸⁹ To move closer to causality, the absolute change in gut microbial species between experimental or phenotypic conditions (eg, lean, obesity) needs to be defined. Absolute measurements would reveal whether the gut microbiota induces an energy deficit that is sufficient to influence weight, and the specific microbial species and functions that are quantitatively important for energy balance. Measurements of absolute microbial abundance and load further allow for the quantification of functional microbial readouts, which are key for uncovering mechanisms by which gut microbes affect energy balance.^40–42,90

There are several methods to achieve absolute quantification of microbial species and microbial load. One method determines the exact number of microbial cells per fecal sample through flow cytometry and discriminates between live and dead cells by staining with propidium monoazide.⁴⁰ To this method, a synthetic DNA spike-in can be added to facilitate absolute quantification of metagenomic reads.⁴¹ Combined, this approach results in an absolute quantification of microbial cells in a sample, resolves species composition, and reveals how many of each microbial species are dead or alive. There are several other approaches for absolute quantification of gut microbes, including digital polymerase chain reaction combined with 16S rRNA amplicon sequencing,⁸⁹ parallelization of amplicon sequencing and flow cytometric enumeration of microbial cells,⁹⁰ and various other spike-in methods.^91,92

Beyond Inferred Function: Community-Wide Functional Readouts

Although gene relative abundance with subsequent inference of function with bioinformatic tools enables determination of the metabolic potential of a microbial community, these genome-centric analyses do not provide a mechanistic understanding of microbial functions associated with health and disease. In addition, individual functions paint an incomplete picture of microbial community dynamics. Direct determination of functions will enable testing of mechanistic hypotheses, design of precise interventions targeting specific microbial functions important to energy balance, and drastically accelerate the translation of findings into the clinic. Achieving this will require an understanding of how microbial communities are assembled, maintained, and function as a system. Uncovering microbe-microbe interactions and how microbes react to perturbations in diet is essential for the development and testing of precision nutrition interventions. This is critical for guiding proactive treatment solutions and targeted microbiome engineering.

One step forward in the area of precision microbiome engineering through diet is the development of ribosome profiling techniques (metatranslatomics) that can elucidate protein translation. This is critical because the presence of a gene or transcript does not provide information on whether a protein is ultimately made.^93,94 In addition, integration of multi-omic data from microbial genomes, transcriptomes, and translatomes advances knowledge by elucidating mechanisms that are not evident with each platform independently.^95,96 Mathematical modeling and machine-learning approaches are essential for interpreting results^95,97,98 and understanding microbial niches⁹⁹ and community dynamics.^100,101

Unraveling complex interactions and correlations of various members of the human gut microbiota can be challenging, given that genome-centric approaches often do not discriminate between DNA obtained from live cells and that from dead cells.^40,102 This is problematic because DNA is a relatively stable molecule and can persist in the environment long beyond the lifespan of a microbe. This limits the elucidation of the functional impact of DNA-based readouts at a particular moment in time. Therefore, measurements of metabolic activity are a vital complement to DNA-based measures and can generate insights into microbial community function and activity.^103,104 Integrated metagenome and metatranscriptome experiments have been performed to identify associations of microorganisms with the onset and progression of disease.^105,106 A recent application of this integrated approach provided insights into the role of the gut microbiota in irritable bowel disease.¹⁰⁷ Moving microbiome science toward an understanding of community-wide functional readouts is a necessary advancement for making the promise of microbiome-directed precision nutrition a reality. Indeed, such approaches are being leveraged to uncover the role of the gut microbiota in precision nutrition interventions including in silico modeling, in vitro methods to identify promising diet interventions that should advance to validation in animal models or clinical trials, and a broad range of in vivo models across the translational spectrum to facilitate causal inference regarding the role of the gut microbiota in conferring the beneficial effects of diet.¹⁰⁸

Innovative, Tightly Controlled, and Comprehensive Clinical Trial Designs

Clinical trial designs represent another area where innovation is needed. Elucidating causal pathways between diet, changes in microbial communities, and obesity requires strategic integration of tightly controlled efficacy trials with deep phenotyping and population-based studies to test effectiveness in real-world settings. One model of such an approach is the ongoing Nutrition for Precision Health study (ClinicalTrials.gov identifier NCT05701657), which includes an observational phase followed by both an outpatient and an inpatient controlled feeding phase. Deep phenotyping will be combined with artificial intelligence and machine learning to stratify dietary recommendations based on an individual’s phenotypic traits, including the composition of the gut microbiota. It is important to state that advancing proof-of-mechanism clinical investigations for generating microbiota-focused dietary recommendations requires iterating back to preclinical studies. Such studies could include cell culture, gut models (eg, gut-on-a-chip,¹⁰⁹ bioreactors^110,111) and mouse models³⁵ to ensure a comprehensive translational pipeline that provides robust evidence of causality.

There are multiple clinical trial design considerations that will advance the development of microbiota-focused dietary interventions targeting energy balance. First, given the reliance of gut microbes on host diet for their energetic needs, all clinical studies of the gut microbiome require an assessment of dietary intake. Validated tools such as food frequency questionnaires or 24-hour dietary recalls should be selected based upon the research question, the population being studied, feasibility, and cost.¹¹² More advanced tools such as ecological momentary assessment and food photography can reduce the biases associated with self-reported diet.¹¹³ Biomarkers of dietary intake continue to emerge,¹¹⁴ with innovations in the use of fecal metagenomic sequences to determine host dietary intake^115,116 opening the door to more precise indicators of nutritional status to complement other diet assessment methods.

With respect to controlled feeding, care should be taken to deliver diets that vary in the specific nutrients hypothesized to affect microbial function, with other nutrients kept as constant as possible. Some study designs (ie, over- or underfeeding paradigms) require varying total energy only, thus making it feasible to standardize macro- and micronutrient composition. Other designs that compare divergent dietary patterns, such as a Western diet and a Mediterranean diet, are going to vary vastly in micronutrients. Thus, the design should maximize the diet drivers (eg, fiber and whole foods) while minimizing differences in micronutrients by including as many similar types of foods as possible. Controlled feeding studies that aim to test energy balance hypotheses should measure not only energy intake but also energy expenditure. An alternative to measuring energy expenditure in large, diverse cohorts is to use validated equations that include variables such as age, sex, body composition (or anthropometrics), and physical activity. Importantly, study diets should be prepared in a metabolic kitchen using research-grade software to accurately deliver nutrients and avoid confounding from unintended over- or underfeeding.⁴³

Because gut microbes are responsive to multiple host factors, well-controlled feeding studies and nutritional challenges performed in a laboratory setting are needed to dampen the influence of environmental factors on the gut microbiota. Crossover studies (when appropriate to the study question) in which each participant serves as their own control are a powerful tool for maximizing the signal to noise ratio in studies using diet approaches to modulate the gut microbiota.¹¹⁷ Studies with parallel-arm designs could benefit from larger sample sizes that allow for the study of interactions between diet and host factors such as age, sex, and body mass index. Combination designs with a controlled feeding component in parallel to an outpatient behavioral or controlled feeding study allow for the assessment of core end points while capitalizing on the unique benefits of each design. This can accelerate translation to clinical practice by providing evidence of efficacy and effectiveness.

Enhancing the precision of energy balance measurements is also essential for studying host response to diet-driven gut microbiome remodeling. Fecal energy assessments may pose the biggest challenge to capturing true interindividual variability as opposed to noise from assay or experimental design limitations. The classic approach of using dyes to mark the start and end of fecal collection is error prone. Methods that combine quantitative markers that achieve a precise readout of the fecal collection time are an advancement on fecal dyes.^14,43 In this area, we recently validated the use of the continuous fecal marker microdosed polyethylene glycol to adjust fecal energy to 24 hours, similar to its prior application as a marker for nutrient balance.^118,119 A second innovation was the implementation of chemical oxygen demand (COD), which provides a more physiologically relevant measure of energy with higher throughput and lower cost.¹²⁰ We previously reported that, for food items, COD highly correlates to bomb calorimetry (R² = 0.97).¹²¹ Coupled with our ability to adjust fecal energy to polyethylene glycol recovery, the use of fecal COD represents a major advancement in human-gut microbiome bioenergetics.¹²²

Finally, deep phenotyping of host and microbes is key. On the host side, clinical variables such as body composition and enteroendocrine hormones influencing appetite and satiety can reveal alterations to body energy stores and ingestive mechanisms that respond to changes in diet and the gut microbiota. Colonic transit time, which influences and is influenced by the gut microbiota, is an essential driver of host-diet-microbe interactions that is variable between individuals and can help explain microbial phenotypes stemming from changes in colonic pH and fermentation.^14,53,56 Studies evaluating energy balance require a comprehensive and precise assessment of the entire energy balance equation. In addition to energy intake, energy expenditure, and fecal energy loss, other variables of interest include substrate oxidation (ie, carbohydrate and fat utilization), metabolic flexibility (ie, adaptation to changes in metabolic fuel availability), and metabolic adaptation (ie, shifts in energy expenditure in response to over- or underfeeding), and energy losses from microbially produced gasses.^{43,75,123,124} Alongside the comprehensive microbiota phenotyping described earlier, novel devices that sample the gut microbiota along the gastrointestinal tract promise to further our understanding of gut microbial ecosystems throughout the small intestine and various segments of the colon. Such approaches also avoid the limitations of fecal sampling, which includes exposure to oxygen, contamination, and variability in results due to sample collection and processing.^125,126 Finally, analytic techniques that appropriately integrate rigorously collected deep phenotypic traits of both the host and gut microbiota will be critical for the success of precision nutrition therapies that target energy balance.

Fecal Microbiota Transplantation

Could FMT optimize energy balance? Unlike in mouse studies, where discordant weight phenotypes can be transferred to the host via FMT,^9,17,127,128 FMT has not been consistently successful in humans.^129,130 This does not discount causality as both host¹³¹ and donor¹³² characteristics affect the efficiency of donor microbiota engraftment and, thus, the transferability of donor characteristics to the host. Of relevance to precision nutrition, interindividual variation in both the composition of host diet and the gut microbiota is likely to affect the efficacy of FMT for weight regulation. This was illustrated in a recent trial in which autologous FMT from samples collected after a polyphenol-enriched Mediterranean diet led to a reduction in weight regain as compared with a generally healthy diet or a Mediterranean diet without polyphenol supplementation.⁵⁰ The success of the green Mediterranean diet was correlated with a maintenance of gut microbiota changes observed after the diet intervention,⁵⁰ including persistent alterations in low abundance (ie, noncore) microbes.^133,134 Importantly, heterogeneity in the underlying pathophysiology leading to obesity may require customized fecal transplants and adjunctive therapies to target different components of the energy balance equation based on an individual’s unique physiology. Overall, harnessing the gut microbiota as a therapeutic strategy via FMT will require a mechanistic understanding of host-diet-gut microbiota interactions and how specific gut microbial community metabolic activities affect energy balance.

Outlook

Leveraging precision nutrition to remodel the gut microbiota for optimizing human energy balance may aid in the prevention of weight gain and promotion of weight loss and weight loss maintenance. Even with provision of highly effective obesity pharmacotherapies, participants demonstrate variability in weight loss response.² Factors such as diet adherence, sex, genetic variation, or other unaccounted-for factors have been suggested as drivers of treatment-response heterogeneity.¹³⁵ Regardless of the true driver of response heterogeneity, microbiota-directed therapies that maximize negative energy balance could complement pharmacotherapies. The implementation of precision nutrition to optimize energy balance through fine-tuning the gut microbiota could align with the “small changes” approach to weight management whereby a small negative energy balance can have a measurable impact on weight over time.^136,137 In addition, small molecules produced by the gut microbiota and enteroendocrine hormones may also reduce appetite and food intake,⁸⁷ which could amplify the effects of the gut microbiota on overall energy balance. Precisely controlled feeding paradigms that reveal the physiological effects of the gut microbiota should be coupled to population-based studies to generate big data that can be mined for stratification of diet- and microbiota-based therapies. The efficacy of FMT could also benefit from precision nutrition approaches that optimize treatments based on host physiology. Together, these strategies may fulfill the promise of precision diets that optimize energy balance through microbiota-directed therapies.

Acknowledgments

The funders had no role in designing the study, the analysis, or the decision to submit the paper. The views expressed are those of the authors and not necessarily those of the National Institutes of Diabetes, Digestive, and Kidney Diseases or the National Institutes of Health.

Author Contributions

K.D.C., D.I., and S.R.S. conceived of the review concept. K.D.C. conducted the literature review and wrote the initial draft of the manuscript. All authors critically revised the manuscript, contributed to the manuscript, and approved the final version.

Funding

K.D.C., S.R.S., and K.Z. are funded by the National Institutes of Diabetes, Digestive, and Kidney Diseases (grant 1R01DK140328).

Conflicts of Interest

The authors have no relevant interests to declare.

References