Allie Johnson, DNM, DIM, PNM

There is a conversation about antibiotics that almost never happens in a doctor's office. It's not about whether you should take them — it's about whether you're already taking them without knowing it, every time you sit down to eat.

The United States Food and Drug Administration reports that roughly 80 percent of all medically important antibiotics sold in this country go not to humans, but to food-producing animals. That statistic has been in the public record for years. It has not changed policy in any meaningful way. And the average person who eats a hamburger, drinks a glass of milk, or orders farmed salmon at a restaurant has never been told.

You can't consent to what you've never been told.

In 2021, the FDA reported that approximately 10.4 million kilograms of medically important antimicrobials were sold for use in food animals in the United States. This number does not include drugs classified as non-medically important — compounds like ionophores, which are used exclusively in veterinary settings and have no human counterpart but still alter the biology of every animal that consumes them, and by extension, the biology of every person who eats those animals.

The rationale for agricultural antibiotic use falls into two broad categories: treatment of sick animals, and subtherapeutic dosing for growth promotion and disease prevention in healthy ones. The second category is where the vast majority of use has historically occurred. Animals in industrial confinement operations — crowded, stressed, often immune-compromised — are routinely given low-dose antibiotics because without them, infectious disease would spread through a facility quickly. The drugs keep the animals alive long enough to reach slaughter weight. They also, conveniently, accelerate weight gain. The mechanism for that growth promotion effect is the same one that concerns every informed practitioner: altered gut microbiome.

The classes used in livestock production read like a catalog of human medicine. Tetracyclines — oxytetracycline, chlortetracycline, doxycycline — account for the largest share. These are the same tetracyclines your doctor prescribes for acne, respiratory infections, and Lyme disease. Macrolides include tylosin and erythromycin, related to the azithromycin (Z-Pack) that has become one of the most prescribed antibiotics in American medicine. Aminoglycosides such as neomycin and gentamicin are given to poultry and swine. Beta-lactams — the penicillin family — are used across species. Fluoroquinolones, including enrofloxacin (the veterinary analogue of ciprofloxacin), were once widely used in poultry before the FDA withdrew approval for that specific use in 2005; they remain in use in other livestock sectors.

Then there is sulfonamide class, which includes sulfamethoxazole — the sulfa half of the widely prescribed trimethoprim-sulfamethoxazole (Bactrim) combination. And colistin, a last-resort antibiotic that physicians turn to when everything else has failed — a drug so toxic to human kidneys that it fell largely out of clinical use decades ago — is still used in livestock production in parts of the world, including countries that export food to the United States.

Ionophores deserve separate attention. Monensin, salinomycin, lasalocid, and narasin are given to cattle and poultry for the dual purpose of preventing coccidiosis (a parasitic gut infection) and improving feed efficiency. These drugs are not used in human medicine at all — they are too toxic for that. The FDA classifies them as non-medically important, which exempts them from the tighter oversight applied to drugs that have human equivalents. There is no withdrawal period requirement meaning animals can be slaughtered while still on the drug. And because they are not tracked under the same reporting systems, the full scale of their use is not publicly quantified the way other classes are.

The assumption that antibiotic residues in food are a meat problem misses a significant portion of the exposure pathway. Dairy cattle receive antibiotics for mastitis treatment — a bacterial infection of the udder that is endemic in factory-farmed herds. Regulatory tolerance levels for residues in milk exist, and bulk tank milk is routinely tested before processing. But the testing is not for every drug. Some veterinary antibiotics are not included in standard screening panels, which means they can and do pass through undetected. Eggs present a similar picture. Laying hens are not supposed to be given antibiotics during laying, but when birds are treated and eggs are inadvertently collected during a withdrawal period, or when small-scale operators don't follow withdrawal schedules precisely, residues appear.

Plant agriculture adds another layer that most people never consider. Streptomycin — the antibiotic most people associate with tuberculosis treatment — is sprayed directly onto apple orchards, pear orchards, and stone fruit crops to control fire blight, a bacterial disease caused by Erwinia amylovora. Oxytetracycline is also registered for this use. Kasugamycin, a broad-spectrum aminoglycoside, has been approved for use on certain crops. These applications are concentrated in the Pacific Northwest, the Great Lakes region, and other major fruit-producing areas. Organic certification prohibits streptomycin on tree fruits in the United States, which makes this one of the clearer cases where the organic label provides a genuine, documented difference in exposure.

Water carries the load that everything else leaves behind. Antibiotic compounds are excreted by animals and humans and enter waterways through agricultural runoff, septic systems, and municipal wastewater treatment plants — which are not designed to remove pharmaceuticals. Studies of US surface water and groundwater consistently detect tetracyclines, sulfonamides, fluoroquinolones, macrolides, and other antibiotic compounds at concentrations ranging from nanograms to micrograms per liter. Municipal water treatment reduces but does not eliminate them. The people drinking from wells near large confined animal feeding operations carry a measurably higher antibiotic load in their water supply.

The regulatory position on antibiotic residues in food is built on a framework of tolerance levels — the concept that a certain amount is acceptable because it falls below a threshold that would cause acute harm. That framing is borrowed from toxicology, and it makes a fundamental error: it treats the human gut as if it is simply a transit system, not an ecosystem.

The human gastrointestinal tract houses somewhere between 38 and 100 trillion microbial organisms — bacteria, archaea, fungi, and viruses — that collectively perform functions no pharmaceutical has ever successfully replicated. They synthesize vitamins, regulate immune responses, produce neurotransmitter precursors, process bile acids, train the mucosal immune system, and maintain the barrier integrity of the gut lining. Disrupting this system at even low levels, repeatedly, over years, is not the same as having no effect. It is simply an effect that takes longer to become visible.

Research published in journals including Nature, Cell Host & Microbe, and Gut has documented that subtherapeutic antibiotic doses — the same doses used in livestock — alter the microbiome composition in animals and humans alike. The changes favor antibiotic-resistant strains. They reduce diversity. They shift the ratio of Firmicutes to Bacteroidetes in ways that parallel the changes seen in obesity, metabolic syndrome, and inflammatory bowel conditions. A microbiome that has been chronically exposed to low-dose antibiotics through food is not a neutral microbiome. It is a shaped one.

Immune dysregulation is the downstream consequence that receives the least attention in public discourse. The gut-associated lymphoid tissue (GALT) is the largest component of the human immune system. Roughly 70 percent of immune cells reside in or adjacent to the gastrointestinal tract. What happens to the microbiome happens to immune surveillance, to oral tolerance, to the calibration of inflammatory responses. The rise in autoimmune conditions, food sensitivities, and atopic disease in industrialized nations tracks closely with the industrialization of food production — including its antibiotic load. That correlation does not establish causation, but it does establish a question that has not been adequately answered.

Antimicrobial resistance is now classified by the World Health Organization as one of the ten greatest global public health threats. The CDC estimates that more than 2.8 million antibiotic-resistant infections occur in the United States each year, killing more than 35,000 people. Globally, the death toll attributed to resistant infections was estimated at 1.27 million in 2019, with projections suggesting it could reach 10 million annually by 2050 if current trajectories continue.

The mechanism of resistance is not complicated. Bacteria evolve under selective pressure. Expose a population of bacteria to antibiotics — even at subtherapeutic doses — and the ones that survive are the ones that carry or develop resistance mechanisms. These mechanisms spread horizontally through a process called conjugation: resistant bacteria can transfer resistance genes directly to other bacteria, even across species. This is not theoretical. The mcr-1 gene, which confers resistance to colistin, was first identified in livestock in China in 2015. Within months of its discovery, it had been found in human clinical isolates in dozens of countries. Within two years, it had reached the United States in both livestock and humans.

Colistin is significant because it is one of the antibiotics of absolute last resort — the drug physicians turn to when a patient has a carbapenem-resistant infection and almost nothing else will work. The idea that a last-resort antibiotic was being used routinely in food production while resistance to it was spreading globally is not a hypothetical risk. It is a documented failure of the regulatory framework that was supposed to prevent exactly this outcome.

Carbapenems present the same pattern. Carbapenem-resistant Enterobacteriaceae (CRE) — a category that includes resistant Klebsiella pneumoniae and E. coli — are now found in both animal and human populations. The genes that confer resistance move between settings. The animal reservoir is a documented source of human infection, and vice versa. The concept of One Health — the recognition that human, animal, and environmental health are inseparable — emerged from this reality. It remains mostly aspirational as a policy framework.

Most conversations about agricultural antibiotics focus on land animals. Aquaculture — the farming of fish, shrimp, and shellfish — receives far less scrutiny, which is a problem because it is one of the fastest-growing food sectors in the world, and its antibiotic use is substantial and largely invisible to the consumer.

Farmed shrimp and fish live in high-density enclosures that are structurally similar to the confined animal feeding operations used for pork and poultry — with the added complication that they are submerged in water, meaning that drugs administered to the animals are also administered directly to the aquatic environment. Tetracyclines, fluoroquinolones, and sulfonamides are all used in aquaculture operations. Some countries where the US imports significant seafood — Vietnam, Thailand, India, Bangladesh, Ecuador — have far weaker regulatory frameworks for antibiotic use in aquaculture than the United States does for domestic production. The FDA tests only a small fraction of imported seafood for drug residues, and the list of drugs tested does not cover all compounds in use.

Studies of retail shrimp purchased in American grocery stores have repeatedly detected antibiotic residues including chloramphenicol (banned from food use in the US), nitrofurans (also banned), and oxytetracycline. A 2015 study published in the Journal of Hazardous Materials found that shrimp imported from Southeast Asia contained antibiotic-resistant bacteria and residues at levels that would not be permitted if the shrimp had been produced domestically. This is not a fringe finding. It has been replicated by multiple independent research groups.

The FDA's framework for managing agricultural antibiotic use shifted significantly in 2017, when the Veterinary Feed Directive took effect and the agency asked the pharmaceutical industry to voluntarily remove growth promotion claims from antibiotic labels — meaning that antibiotics marketed specifically for accelerating weight gain could no longer be sold as such. This was presented as a meaningful reform. In practice, it was the equivalent of asking tobacco companies to stop marketing cigarettes as slimming, without restricting the sale of cigarettes.

The same drugs, at the same doses, can still be administered under the label of "disease prevention" rather than growth promotion. The distinction is paperwork, not pharmacology. A veterinarian's signature is now required — but in a system where a single accredited veterinarian may oversee hundreds of thousands of animals across multiple facilities, the oversight this implies is largely administrative. The FDA's guidance documents on "judicious use" are exactly that: guidance. They are not enforceable regulations. They do not cap the total volume of antibiotics that can be used. They do not require public disclosure of what is used and at what doses, beyond the aggregate national sales data published annually by the FDA.

There is no comprehensive mandatory residue testing program that covers all approved veterinary drugs across all meat and poultry sold in the United States. The USDA's National Residue Program tests for some compounds in some species at some slaughter facilities. The gaps are systematic, not incidental. Ionophores, for instance, are not routinely tested for because they are classified as non-medically important — despite the fact that they are biologically active compounds that alter the microbiome of every animal that consumes them and that no withdrawal period is required before slaughter.

The informed consent framework that governs human medicine — the principle that a patient has the right to know what substances are entering their body and to consent or decline — does not extend to the food supply. The antibiotic residues, the resistance genes carried by bacteria on food surfaces, the ionophores in conventionally raised beef: none of these appear on any label. Nobody asks your permission. Nobody tells you.