Calorie Counts Are Wrong: Gut Bacteria Absorb Hundreds More Per Day

The number printed on a nutrition label has been calculated using the same formula since the 1800s, and it has a significant blind spot: the trillions of bacteria living in your colon. A new mathematical model developed by researchers at Arizona State University now accounts for that microbial layer, and the results published in PLOS ONE suggest that two people eating identical meals can absorb calories that differ by hundreds per day, depending on their gut microbiome.

DAMM Model Outperforms Century-Old Atwater Calorie Formula

The Atwater factors, the standard method for estimating how much energy humans extract from food, were developed in the late 19th century. They assign fixed absorption coefficients to protein, fat, and carbohydrate, but they treat the large intestine largely as a passive tube. Taylor L. Davis and colleagues at Arizona State's Biodesign Center for Health through Microbiomes built the Digestion, Absorption, and Microbial Metabolism model, or DAMM, to fix that omission.

The improvement in predictive accuracy was substantial. When tested against clinical trial data from two distinct diets, the DAMM model raised the R² fit from 88% under the Atwater approach to 96%. More telling, it slashed the mean difference between predicted and measured metabolizable energy from -22.3 gCOD per day to just -2.5 gCOD per day, eliminating most of the systematic bias that Atwater-based calculations carry.

How Gut Fermentation Adds and Subtracts Usable Calories

The mechanism behind DAMM's improvement comes down to microbial fermentation in the large intestine. When fiber and other undigested material reach the colon, resident bacteria ferment it in a process that functions, as the researchers put it, like a microscopic brewery. That fermentation produces short-chain fatty acids (SCFAs): acetate, propionate, and butyrate. These molecules pass directly through the colon wall into the bloodstream, where the body burns them for energy. Butyrate in particular feeds the cells lining the colon, reduces inflammation, and has been linked to protection against colon cancer.

The bacteria, however, do not always add calories back to the host. Some gut residents, called methanogens, convert hydrogen produced during fermentation into methane gas, which is exhaled. That represents energy leaving the body entirely. DAMM tracks both directions: SCFAs gained, methane lost. The model validated that accounting by measuring total methane released by clinical trial participants over 23 hours inside a whole-room indirect calorimeter.

This dual accounting matters because the gut microbiome regulates energy balance through mechanisms beyond simple fermentation, influencing metabolic genes and fat storage in ways that fixed calorie tables cannot capture. The DAMM model gives researchers a quantitative handle on the fermentation side of that equation for the first time.

Clinical Trial Validation Across Two Distinct Diets

To test whether DAMM held up against real human data, the team validated it against a registered clinical trial (NCT02939703) that put participants on two deliberately different diets. The model had to account for variation in macronutrient composition, microbial substrate availability, and individual differences in methane output. It captured both the direction and magnitude of change in measured metabolizable energy across both dietary conditions, something the Atwater formula routinely failed to do.

The mathematical structure of DAMM breaks diet composition into the gross energy of individual macronutrients, calculates direct absorption in the upper gastrointestinal tract, and then applies microbial stoichiometry to estimate what bacteria consume from the residue that reaches the large intestine. That stoichiometric step is what quantitatively predicts SCFA and methane output rather than ignoring them. The clinical validation is a meaningful threshold: gut microbiota animal studies have long shown microbial influence on weight, but translating that to a predictive human model has been elusive, and human study results have historically been inconsistent.

What the Model Cannot Yet Prove About Obesity

DAMM is a tool for measurement, not a proof of causation. The study makes clear that the model can show how bacteria affect calorie absorption in a given individual on a given diet, but it does not establish that any particular microbial community causes obesity. That distinction matters because gut microbiota human study inconsistencies have frustrated attempts to translate animal findings directly into human weight management strategies.

The model also does not yet capture the full complexity of microbial metabolism: interactions between microbial species, host immune responses, and the time-dependent dynamics of how diets shift microbial communities over weeks and months are not baked in. Davis described DAMM explicitly as a framework designed to evolve, one that can incorporate new insights on diet, metabolism, and microbial interaction as the field advances.

Personalized Diets and Metabolic Disease Research Are Next

The near-term applications point in two directions. For researchers, DAMM can be linked to existing human body composition models to predict how changes in gut microbial activity ripple through energy stores over time, offering a more mechanistic window into obesity, type 2 diabetes, and other metabolic disorders. More than two in five U.S. adults have obesity, with downstream risks for cardiovascular disease and certain cancers, so better models of energy balance carry real clinical weight.

For clinicians, the longer-range goal is using a validated DAMM-style framework to design personalized diets based on an individual's actual microbial profile rather than generic label calories. That would mean pairing a dietary prescription with microbiome sequencing and using the model to predict how many calories a specific person will actually absorb from a specific eating plan. The researchers see emerging interventions including precision microbiome therapies such as engineered probiotics and targeted prebiotics as future inputs the model could eventually incorporate.

For now, DAMM represents the most quantitatively rigorous attempt yet to put a number on what the microbiome actually contributes to human energy balance, and its 96% predictive accuracy suggests that number is worth knowing.