Phages as next-gen human weight modulators
As we resolve the microbes driving obesity, phages could help us debug.
Disclaimer: Phage biology is a rapidly evolving field. All the data and literature I present in this piece are current as of Nov 10, 2025. If you’re reading from the far future, I always recommend reading up on the latest primary research.
Table of Contents
The Curious Case of the Growing Cows
Small Cells, Big Weight
The Silent Sparkplug of Disease
Phages to Debug the Gut Microbiome
The Microbial Targets of Obesity
The Math of Viral Weight Modulation
A Phage’s Purpose
The Curious Case of the Growing Cows
Most of our planet’s antibiotics don’t end up in humans, but in cows.
In the 1950s, microbiologists converged on a strange finding. When sick cows were fed with the antibiotic aureomycin, it didn’t just clear their infections. It made them….fatter. Soon after, they observed the same effect in healthy dairy cows, with some gaining up to 5% of their bodyweight — or several tens of pounds — in a matter of weeks (Marshall et al., 1957).
The aftermath isn’t hard to imagine. Word spread. Excitement boomed. Ranchers around the world pounced on the idea, and we’ve been fattening cattle with antibiotics ever since.
At least, that’s the story we hear.
What we ignore is the mystery buried in plain sight — one that has tortured scientists for decades. Why do antibiotics make cows fatter at all?
Small Cells, Big Weight
In the realm of microbiology, the connection between weight and the microbiome has been percolating for years.
Metagenomic surveys of lean and obese populations showed directional shifts in microbial composition and diversity (Aoun et al., 2020). With time, we found that specific genera in the gut microbiome could manipulate nutrient absorption, fat storage, insulin sensitivity, and global metabolism (Van Hul et al., 2023). And we showed that perturbing these microbial communities ( i.e. via antibiotics) could reliably affect weight, which could be restored with bacterial re-transplantation (Zhang et al., 2019).
Microbiology offers some insight into our mystery. The weight gain we see in cows is less likely the result of antibiotics, and more likely a symptom of carpet-bombing the microbiome.
While the execution is unfortunate, the results are undeniable. There is a powerful interface between weight and the microbiome. And if we could engineer it— without the collateral damage of antibiotics — we have a straight shot at tuning human weight.
Fortunately, nature is home to a far more elegant solution. Over billions of years, bacteriophages have learned to infect bacteria with freakish precision. Because of this superpower, much of our field has focused on turning phages into antibiotics. Meanwhile, we’re ignoring how the same precision could let phages engineer the human microbiome.
As we continue to reformulate disease through the lens of microbial disarray, phages could be the ultimate tool to restore health through the microbiome. If microbes are as crucial to human physiology as we think, phages could be poised to flip medicine on its head.
We’re approaching a future where every dysfunction in human health could be targeted with phages. And obesity happens to be our most promising target.
The Silent Sparkplug of Disease
Why target obesity, of all conditions?
For one, it’s upstream of almost every disease that matters to us. Cancer. COVID-19. Type 2 Diabetes. Alzheimers. Atherosclerosis. It’s hard to find a single chronic health condition that isn’t driven or worsened by obesity.
This is precisely what makes it important. Obesity is perhaps our closest example of a meta-disease—a single target that lies upstream of almost every single condition that plagues humanity.
It’s well-known that most cases of obesity can be managed with lifestyle. But while lifestyle changes are effective at the individual level, they’re proven to fail at societal scales.
As I write this article, nearly every nation with a GDP per capita ≥$10,000/person/year has an obesity rate above 10% (NCD-RisC, 2024)1. While this is a simplification, the trend is unmistakable. With economic prosperity comes an explosion of processed, calorie-dense food and an increasingly sedentary lifestyle. And the result is unprecedented obesity.
As long as we live in a free and wealthy society, obesity will persist. And no clear-eyed person is suggesting we abandon free and wealthy society.
If we believe that obesity is inevitable, there is only one question that matters. Do we choose to prevent it? Or do we scramble to treat it when it takes hold and shapeshifts into every other disease imaginable?
Because it seems like our current solutions are doing the latter. Today’s methods to manage obesity are remarkably tasteless. Until the release of Wegovy in 2021, our options were limited to dietary changes, bariatric surgery, and a handful of prescription drugs with questionable safety profiles (Sherman et al., 2016).
A glimpse of this data speaks to our reality. Most of these approaches are prohibitively expensive, riddled with off-target effects, and still fail to produce sustainable weight loss. And all of them are reactive patches to obesity.
Meanwhile, obesity has the strongest microbial argument of any disease.
Our first glimpse into the microbe-metabolic axis2 emerged from Fredrik Bäckhed’s landmark studies on microbial transplants. Nearly two decades ago, Bäckhed’s group found that germ-free mice transplanted with microbes from obese donors deposited significantly more fat than recipients from lean donors (Bäckhed et al., 2004).
With time, we solidified this phenomenon from several angles.
Subsequent studies found that the ratios of two gut genera (Firmicutes and Bacteriodes) allowed mice to gain less weight than conventional mice with an identical diet (Murphy et al., 2010). Biochemically, we mapped dozens of microbial pathways that could alter weight—from lipid oxidation to inflammation to hunger signaling (Blaser et al., 2013; Carmody et al., 2023).
The science isn’t controversial anymore: our microbiome plays a key role in tuning our weight. What if we could engineer it?
Phages to Debug the Gut Microbiome
Using antibiotics to tune the microbiome is like knitting a sweater with a baseball bat.
The central bottleneck is precision. Any effect on weight comes at the cost of upending the entire microbiome, often for months (Ramirez et al., 2020). It’s no wonder that cattle raised on subclinical antibiotics develop more opportunistic and drug-resistant infections than almost any other animal (Trevisi et al., 2014).
Phages, on the other hand, are ferociously specific. Since 1950s, we’ve known that phages can distinguish between single-receptor variants of bacteria (Abedon et al., 2010). Pharmacokinetic studies showed that certain phages could be cleared from the body in minutes, meaning their effect on the microbiome is likely transient and reversible (Dąbrowska, 2019).
While these quirks make phages tricky as therapeutics, it makes them perfect to engineer the microbiome.
On paper, phages are the perfect vehicles to modulate human weight. This motivates another question: What’s stopping us?
In practice, it’s not very useful to enumerate the benefits of phages.
If the approach was convincing enough, it would already exist. What’s far more useful is inverting the problem. What are the key technical challenges preventing phages from engineering the microbiome?
If we’re rigorous, we’ll find that our idea assumes the following:
Isolate a bacterial strain that consistently and mechanistically drives obesity
Discover + engineer phages that specifically lyse this bacteria
Efficiently deliver these phages to the gut for transient suppression in vivo
To paraphrase: we need a sharp weapon, and a reliable target.
Fortunately, we’re not starting from scratch.
It turns out that phages have been formulated as oral therapies for nearly a century, ever since Félix D’Herelle used them to treat the first cases of bacterial dysentery (Summers, 2012). After years of optimization — especially in phage centers like Eliava — we’ve developed a solid foundation of in vivo data, and a significant body of research on enteric phage delivery3.
While these challenges certainly aren’t trivial, I’ve already touched on these in my piece on phage immunogenicity, and covering it any fairness would be another article. As the literature suggests, the packaging of our phages is likely the most stable of our assumptions.
If anything, the largest uncertainty of engineering the microbiome doesn’t lie in our phages, but in our ability to find and engineer bacteria with precision.
Let’s dissect this carefully.
The Microbial Targets of Obesity
Which bacteria has the strongest implication in human obesity?
After careful analysis of the research, I’ve realized that this is an open question.
Longitudinal genomics reveals that the gut microbiome is remarkably plastic — shifting with diet, geography, and environmental stimuli — even within the same individual over time (Xu et al, 2014; Yatsunenko et al., 2012).
It follows how this complicates our solution. Even if we found a bacterial species associated with obesity, it’s a real task to find a baseline level of bacteria associated with sickness or health4.
And even if we could catalog all the bacteria in the gut, they would still be troublesome to study. How can we tell, for instance, if a bacterium drives obesity or simply thrives on a high-sugar diet that drives fat gain?
Testing this involves knocking specific strains of bacteria in and out of the gut, which we can only partially achieve with antibiotics or microbial transplants5. While this doesn’t invalidate our answers, it certainly limits their resolution.
But what does the literature say?
Perhaps our strongest evidence for obesogenic bacteria leads us to a species called R. gnavus. It’s a gram-positive anaerobe thought to modulate weight by extracting calories from glycans that line the intestinal lumen.
Metagenomic surveys show that R. gnavus is a universal resident of the gut microbiome, and has been isolated from humans irrespective of geography, diet, and environment (Nooij et al., 2025). Population-level studies have also reveal a positive relation between R gnavus and obesity, and levels of R gnavus are predictive of weight gain patients receiving fecal microbial transplants (Palmas et al., 2021; Hong et al., 2023).
While this is great, identifying a bacterial culprit feels shallow. If we designed a therapeutic, what phages would we need? How much? How long would the dose last? How would we avoid resistance or downstream effects on microbiome?
Sculpting the gut microbiome isn’t as simple as cherrypicking bacteria and phages. We need to understand the mechanics.
The Math of Viral Weight Modulation
Again, isolating bacterial targets against obesity is no small feat. But for the sake of a model, let’s pretend our target was R. gnavus. How might we use phages to sculpt this bacterial population and control weight?
The first step, of course, is finding a lytic phage that’s active against our target. A quick search on NCBI shows that we’ve already isolated and sequenced at least ten phages against R. gnavus. One workable example is phage RgPS6, which is known to lyse R. gnavus without infecting other bacteria in the same genus6.
After identifying a phage, we can use sequential evolution to optimize its virulence in the gut microenvironment7. We can then propagate the phage to billions of pfu/mL and measure its ability to suppress bacterial growth.
Which leads us to another relevant question—exactly how many phages would we need?
While questions like these can only be resolved experimentally, I decided to crunch the numbers.
Last weekend, I sketched out some differential equations to simulate phage dynamics in the gut. For context, I structured my math on the Levin–Weitz–Beretta regime of viral kinetics, with some fine-tuning to capture imperfect delivery and phage decay in the human gut (Galtier et al., 2016).
For context, B(t) is the population of uninfected R. gnavus population (cells/mL). I(t) is the population of infected R. gnavus cells undergoing lytic replication (cells/mL). P(t) is the concentration of free bacteriophages in the gut lumen (PFU/mL):
After finding some scientifically-grounded values for each of these variables, I modeled the behavior of a one-shot oral phage dose in MATLAB. Specifically, I wanted to solve for how many phages we’d need to kill 50% of R. gnavus in the gut in 24 hours:
This reveals some valuable insights. Even if we were reckless enough to rapidly eliminate half the R. gnavus in our gut, we’d need a one-shot dose of ~6x10¹¹ plaque-forming units. Once again, mathematics can’t substitute for a real experiment — but it does offer an order of magnitude to ground our thinking.
Even if we massively underestimated our phage dose by 100x (i.e. improper delivery, chemical degradation), this figure still falls well within the range of propagation. With some optimization, it’s even plausible to squeeze a working dose of phages into a 1mL, weekly oral pill.
There’s another observation worth noting. Our graph for bacterial load shows that our population of R. gnavus rebounds in a matter of days, as the bacterial community turns over and phages are cleared from the gut. Interestingly, these timescales align with our literature on phage decay in vivo. And it suggests that we could modulate human weight with minimal long-term damage to the microbiome8.
To any phage researchers reading this, it’s obvious that none of these steps are magical. All of these protocols fall within the realm of basic phage biology. These are experiments we can run today. And they are simple enough to produce meaningful answers in a month’s time.
Still, hardly anyone is trying to engineer phages to control weight. For all its potential impact, the concept is seriously overlooked.
In other words, it’s wide open for research.
A Phage’s Purpose
The next few years could be the most exciting season in phage research.
Since the invention of germ theory, bacteria were shackled to the world of infectious disease. Phages were limited to basic science and seen as unwieldy therapeutics.
But the truth is stirring. In recent years, microbes have been reported in every major organ system — manipulating processes in innate immunity, nutrient assimilation, endocrine function, and cell signaling. Diseases that we long assumed to be purely genetic and environmental — cancer, diabetes, Alzheimer’s — are being revisited from a microbial perspective.
Microbiology could be our most flagrant gap in modern medicine. And phages are poised to become the foundation of microbial therapeutics.
We can debate the speed, but the trajectory is inevitable.
Obesity is arguably the largest meta-problem in health, and it could be the most immediate application of phages. In a weight loss market bloated with subpar solutions, phages have all the markings to become a cheaper, safer, sustainable alternative. And developing them is already in the range of today’s microbiology.
This advance alone could make for a life’s work. But it points to something far greater.
As we discover that bacteria run our world, phages are poised to become unfathomably useful. We’re approaching a future where microbes are sculpted in service of humanity. Phages could become the chisel that lets us reshape biology itself. And as phage researchers, we’re in the ideal position to help.
This is the true potential of microbiology — it’s the next logical step in the evolution of phages.
Let’s be remembered as the people who made it happen.
Footnotes
Some notable exceptions include Japan and South Korea, which offers some fascinating lessons and warnings to the rest of the world (in terms of diet, culture, economics, city design, etc.). Even so, neither nation has been able to prevent obesity rates from rising in the past decade.
Here’s an insightful meta-analysis from Nature from on how infrastructural choices (like the walkability of a city) can affect health outcomes of its citizens.
I think I made up this term
For context, phage centers like Eliava have spent years administering their “pyophage” cocktail for enteric infections.
Meanwhile, in the world of basic science, we’re attempting to package oral phage therapies with every delivery mechanism imaginable — from hydrogels to LNPs to pH-dissolving capsules.
For context, consider how the majority of illnesses (i.e. cancer, hypertension, diabetes) are either linked to abnormal biomarkers or highly specific physiological traits.
It’s a trifling observation, but this is the foundation of how we diagnose and treat disease. If we can’t find a cutoff between healthy and non-healthy levels of gut bacteria, it certainly adds work to our end.
This is another use-case where phages can be immediately valuable. Since basic research on the gut microbiome is still bottlenecked by precise bacterial control, bacterial viruses are already in position to help.
The current consensus is that RgPS6 is a temperate (lysogenic) phage. That said, it was still shown to have measurable lytic activity against R. gnavus (which is rarer, but known to be possible).
One promising alternative is the MOR family of phages isolated in 2023 against R. gnavus CC55_001C.
Beyond improving our phage for in vivo use, this also lets us modify its genome enough to count as a distinct viral species (and therefore be patentable). As trivial as this sounds, intellectual property has been a longtime speed bump in phage research. No patents mean no money :(
At least, not to the extent that we see with broad-spectrum antibiotics.