Antibiotics in Food: Can Less Do More?

As policymakers push ahead, the data lag behind. "Generally speaking we don't know a lot about how antibiotics are used in animal agriculture," Scott said. In the US, the Animal Drug User Fee Act requires farmers to report their antibiotic use, and classifies the drugs based on their importance in human medicine. Those reports provide an overview of the quantities of antimicrobials used in agriculture, but don't detail why they were used. Meanwhile, regulators have focused on limiting the use of "critical" antibiotics, meaning those deemed most important in the clinic, but different regulatory bodies identify different drugs as critical.

Scott argues that the field needs help from social scientists to examine producers' motivations and behaviors around antibiotic use. Surveys of the animal agriculture industry reveal, for example, that veterinarians and farmers generally feel a moral obligation to treat ill animals. However, the two groups part ways on prevention, with most farmers saying that giving antibiotics to healthy animals to prevent disease is appropriate, and most veterinarians disagreeing. "We need to understand what some consider sacred and some consider profane in avoiding protracted conflict," said Scott.

Rich farm, poor farm

Delia Grace discussed the role of antimicrobials in agriculture in developing countries, which are by far the biggest—and most poorly studied—consumers of these drugs. Grace began with an overview of trends in the livestock industry worldwide, where skyrocketing demand for meat is driving huge increases in animal production. "Demand-driven things are much harder to change than supply-driven things," Grace said.

Indeed, the momentum of meat demand seems unstoppable. A staggering 98% of Earth's mammalian biomass now consists of humans and domesticated animals. People are moving out of poverty at accelerating rates, correlating perfectly with rising demand for animal protein. The richer people get, the more meat they want.

To feed this expanding appetite, farmers are turning to progressively more intensive animal husbandry and raising genetically inbred herds and flocks selected for maximum yields. "When you get animals which are very genetically similar, infection spreads very, very rapidly through them," Grace said. Big outbreaks and chronic infections in enormous animal facilities inevitably call for large amounts of antibiotics.

As global wealth rises, so does meat consumption.

Most developing countries keep poor or no records of animal antibiotic use. To better understand the trends, Grace and her colleagues constructed a mathematical model based on countries' gross domestic products, which track closely with meat demand and the proportion of intensively kept animals, a term that covers all types of large-scale feeding and rearing operations. Assuming that intensive farms in low-income countries use antibiotics at rates similar to those in high-income countries, the investigators estimated a global demand of about 65,000 tons of antibiotics annually for animals. That compares to a global demand of about 100,000 tons for humans.

Looking at actual drug use in specific countries, however, reveals the model's limitations, suggesting that different antimicrobials are used in low-income countries. In Uganda, farmers observed that AIDS led to weight loss, and reasoned that the drugs prescribed to treat HIV would therefore cause weight gain. When the government began giving away antiretrovirals for free, farmers started feeding the compounds to their pigs. Meanwhile, Vietnam imports 6,000 kilograms of the asthma medication clenbuterol annually, but only 10 kilograms of it is used medically; the rest is fed to farm animals to increase lean meat production. In contrast, smallholder farms that support single or small groups of families in some parts of Africa suffer extensive animal losses from preventable and treatable diseases, because farmers can't get necessary antimicrobials. Depending on the country, farmers are likely using either too little medication or too much in their animals.

Grace cautioned that a single solution, such as a ban on antibiotics as growth promoters, won't fit all parts of the world: "It's a Goldilocks challenge if you like, it's not a linear problem, we need to address the 'too little' problem as well as the 'too much,' and our normal policy is very bad at that."

Dutch treatment

Jaap A. Wagenaar began his presentation by asking the audience one of the central questions behind the conference: to what extent does antimicrobial use in animal agriculture drive resistance in medically important bacteria? Wagenaar then explained that the answer probably varies dramatically for different antibiotics and bacteria. For example, Campylobacter infections in humans that are resistant to fluoroquinolone antibiotics are quite likely to have originated on a farm, while resistant Neisseria gonorrhoeae infections are highly unlikely to be related to antibiotic use in agriculture.

The effects of policy changes are similarly hard to pin down. The Netherlands banned the use of antibiotics as growth promoters in 2006, but Wagenaar explained that therapeutic use of the drugs on farms shot up immediately afterward. That led to an overall increase in animal antimicrobial use in the years following the ban.

A survey by the European Union in 2009 and 2010 revealed that the Netherlands had unusually high rates of animal antibiotic use, prompting widespread public concern. The news dovetailed with earlier reports of livestock-associated methicillin-resistant Staphylococcus aureus (MRSA) and an outbreak of Q fever that prompted the nation to slaughter all of its pregnant goats. "It was a clear signal that we had to do something on the veterinary side," Wagenaar said.

In 2010, the Netherlands established the Veterinary Medicines Authority agency, which tracks, analyzes, and publishes data from all 72,000 farms in the country, and defines benchmarks for antibiotic use by species. The Authority calculates the amount of antibiotics farmers are using in proportion to their herd sizes, and publishes the numbers. When a farm exceeds the benchmark for drug usage, purchasers voluntarily stop buying its products because they fear public rejection of them. "That economic driver works extremely well," Wagenaar said.

Looking at the results of the new system reveals that veterinary antimicrobial sales have declined by over 58% since 2009; importantly, fluoroquinolones and third and fourth generation cephalosporins—often considered among the most critical antibiotics in humans—are now reduced to minimal use in animals. At the same time, rates of antimicrobial resistance have declined somewhat in broiler chickens, pigs, veal, and dairy cows. A report released in March 2016 by the Veterinary Medicines Authority concluded that the new system has not harmed animal welfare.

Wagenaar credits the availability of good farm-level tracking systems for the country's success in reducing antimicrobial use. "When you do not have data, you have no idea what you're working with, but we had data on resistance and usage," because of the 2010 policy change, said Wagenaar.

Known knowns and known unknowns

After the first three presentations of the conference, the speakers returned to the stage for an extended Q&A with the audience. The group talked about the current public pressure for policy changes, especially in rich countries with comparatively minor food safety and antibiotic stewardship concerns. Grace reiterated her concern that these "worried well" nations might try to impose policies with unintended consequences on poor countries.

But regardless of what wealthy countries such as the Netherlands and the US do, speakers agreed that it will make little difference in the global use of antimicrobials. Developing countries already consume the bulk of antibiotics, and that proportion is rising as they eat more meat.

Responding to a question about the rates of clinical antimicrobial resistance, Wagenaar conceded that his group has not yet seen evidence that reducing the use of these drugs in animals has had an effect in humans.

Audience members also asked about research priorities in agriculture, such as developing vaccines, testing alternative husbandry practices, and identifying non-antimicrobial compounds that might have the same growth-promoting effects without driving antibiotic resistance. While the speakers generally supported such efforts, they pointed out that funding for research in general has not kept up with the need for new innovations.

Eliminating routine antibiotic use could also affect food safety. Mathew described a study comparing conventional and drug-free production of broiler chickens, young chickens that are raised for meat. The drug-free regimen substituted antimicrobials with essential oils, vaccines, and optimized animal husbandry. At the end of the growth cycle, the drug-free chickens were smaller, and were more likely to have Clostridium perfringens, a cause of serious foodborne illness, in their barns. "So we may have a food safety risk here to deal with [when] we take some of these promoters out of the feed," Mathew said.

Farmers who've eliminated antibiotics as growth promoters often rely on selling their products at a higher profit to distributors willing to pay a premium for "antibiotic-free" meats, eggs, or milk. In the case of a disease outbreak, these farmers might treat an entire herd with drugs, then sell that batch of products to a less choosy distributor at a lower price. Mathew argues that this may create an incentive to either withhold treatment from sick animals or not be fully truthful about the farm’s drug use. On a national scale, the farms most able to adopt drug-free practices will likely be the larger ones, perhaps forcing many smaller operations out of business.

Udder resistance

David R. Wolfgang described the current, and coming changes to, applications of antibiotics on dairy farms. Unlike broiler chickens or beef cattle, which follow short, linear paths from birth to slaughter, dairy cows tend to remain on a farm for years, producing milk and consuming feed and antibiotics. Besides milk, one of the major byproducts is manure. Wolfgang explained that "in that [manure] are lots of antibiotics" and bacteria. The impact of that steady stream of antibiotics, however, is unclear.

On dairy farms, 60% of antibiotics are used to treat mastitis, usually by intramammary injection. Because dairy cattle enter milking parlors daily, they're relatively easy to medicate with high concentrations of short-acting antibiotics. In contrast, beef cattle spend much of their lives grazing on range land before being captured and moved to a finishing lot, so medication regimens for beeves tend to be longer-acting, with the active compounds present in the animal at lower doses. The two different protocols apply different selective pressures for antimicrobial resistance.

Studying a single dairy herd for several years, Wolfgang and his colleagues found a few bacterial clones that tended to dominate and persist in milk and mastitis-afflicted udders. The clones possessed some resistance genes for antibiotics that were never used on the farms, possibly highlighting the co-selection of resistance genes on plasmids or other mobile genetic elements.

Salmonella bacteria are very hard to eliminate from a dairy herd.

Efforts to eradicate Salmonella spp. from the herd that Wolfgang's group studied failed consistently. "Some cows have Salmonella almost all the time, and there's a small group ... that never seems to have a positive [Salmonella test]," Wolfgang said. Drug resistance profiles showed a similarly broad distribution, with some resistance genes showing up commonly while others were completely absent; the resistance genes didn't correlate strongly with the antibiotics used on the farm.

Looking at a larger group of farms around Pennsylvania, Wolfgang and his colleagues found diverse bacterial colonization and antibiotic resistance patterns. "One of the challenges I think is when we take samples on farms, if you pick the right day and the wrong group of animals, you may get a completely different image of the resistance problems on that farm," Wolfgang said.

When a dairy cow stops producing milk, it usually gets sold to a beef packing plant. These cull dairy cows arrive at the slaughterhouse in small groups from across the country, and their diverse bacteria spread freely to other cattle at the plant. Wolfgang argues that changes in antibiotic stewardship at the farm level will mean little if they're not accompanied by more careful monitoring of the rest of the food supply chain.

Northern exposure

Agnes C. Agunos presented data on Canada's efforts to reduce cephalosporin antibiotic use in broiler chickens. Third-generation cephalosporins are particularly important antibiotics in humans. Chicken farmers have also used similar drugs such as ceftiofurto treat infections in young chicks, but a 2003 study revealed that cephalosporin-resistant bacteria can persist and contaminate meat sold in supermarkets and butcher shops. A follow-up study in chicken hatcheries in Québec, a province recognized with high agricultural usage rates of these antibiotics, found all of the flocks carried at least one strain of cephalosporin-resistant bacteria. In Ontario, where farmers used these drugs much less, resistance was far less common.

The studies were among the first projects by the Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS), within the Public Health Agency of Canada. CIPARS collects data on antibiotic resistance from farms, food processors, and retail food markets as well as medical clinics and hospitals. The integrated information allows scientists and regulators to identify trends in resistance among people and food animals.

With the baseline rates of cephalosporin resistance in chickens in Québec and Ontario established, CIPARS investigators next turned their attention to human data. Comparing the rates of cephalosporin resistance in Salmonella Heidelberg from chicken hatcheries and medical clinics over time revealed a troubling correlation: "Resistant isolates in chickens and humans are moving toward the same direction, however the association is much more pronounced or stronger in Québec," Agunos said. A voluntary withdrawal of the drugs from Québécois chicken farms correlated with a small but noticeable decline in drug resistance rates in bacteria from both chickens and humans.

In response to public pressure, the Canadian poultry industry voluntarily changed its animal husbandry guidelines in 2014, banning prophylactic use of antimicrobials identified as critical for humans. That group of compounds includes cephalosporins. The change prompted an immediate reduction in the use of cephalosporins, and CIPARS has found a decrease in cephalosporin resistance among bacterial isolates from some farms and retail stores. However, overall cephalosporin resistance in human Salmonella infections in Québec and Ontario has held steady. "This intervention has highlighted that in certain situations, reduction in resistance can happen quickly after the introduction of a use-reduction policy, and there are multiple impacts to measure," including rates of clinical infection with resistant bacteria, said Agunos.

Dismal dosage

Stacy E. Sneeringer and her colleagues study the economic effects of antibiotics in agriculture, with an eye toward predicting the financial impact of changing US regulations. While all meats, eggs, and dairy products sold in the US should already be free of antibiotic compounds, regulators are now focusing on the antibiotic-resistant bacteria that can persist on animal products after livestock have been treated. To combat this persistence, regulators and producers in the US are now phasing out the use of medically important antimicrobials for growth promotion and, in certain states, routine disease prevention.

The structure of modern American farming makes it hard to measure the extent of antibiotic use. In the broiler chicken industry, for example, big producers typically maintain contracts with independent farmers; the producer provides young animals and feed, and the farmer raises the animals. "What that means is that the contractor knows what's in the feed, but the farmer may not," Sneeringer said. As a result, surveys of farmers yield ambiguous data, with large proportions of respondents indicating they don't know whether they use antibiotics, and if so, whether they are used for growth promotion or disease prevention.

Many chicken farmers don't know whether they use antibiotics for growth promotion.

Nonetheless, the survey found that a large percentage of both chicken and hog producers already restrict their antimicrobial use to treatment. That may be because the growth-promoting effects of antibiotics have faded over time. Before the 1980s, adding antimicrobials to feed routinely yielded double-digit percentage increases in animal weight gain and feed efficiency, but over the last twenty years that figure has dropped to as low as one percent in some operations. Sneeringer hypothesizes that this decrease reflects improvements in animal husbandry, such that the antibiotics are no longer necessary for the animals to maintain peak health. "The industry has evolved. To the extent that the growth-promoting antibiotics were serving in a disease-preventing [role], that's no longer having the same effect as it used to," she explained.

Combining the survey results with the latest data on the impact of antimicrobial growth promotion, Sneeringer's team built a model to predict the effects of phasing out routine antibiotic use on chicken and hog farms. The results are reassuring, suggesting that if the underlying data are correct, withdrawing antibiotics from the feed supply should have little or no effect on the overall revenues and meat prices. "What we see is ... very limited impacts on prices and quantities," said Sneeringer. Because these are overall effects, they mask what may happen to individual farmers, like those of concern to Mathew.

Pushing and pulling against resistance

William J. Hall continued the economic theme with a look at the broader impact of antimicrobial resistance. Hall and a team of other economists and policy experts analyzed current trends, and determined that by 2050 antimicrobial-resistant infections could be a bigger cause of global deaths than cancer is today. Through this massive loss of life, plus medical costs and lost productivity, treatment-resistant microbes could drain as many as 100 trillion dollars from the global economy over the next 35 years. "The economic cost is huge, and we think this presents a stark case for action," Hall said.

To find ways to avoid that terrible cost, Hall and his colleagues looked at the underlying causes of antimicrobial resistance. Overuse of antibiotics in both humans and agriculture, and releases of drugs into the environment from sewage and pharmaceutical plants all appear to contribute to the problem, but researchers are still trying to determine which of those categories poses the biggest threat.

Meanwhile, the pharmaceutical industry's development of new antibiotics has faltered in recent years. "There's a clear market failure here, where in a very basic sense companies can make more money by investing in other areas" besides antimicrobials, Hall explained. To address the need for more antibiotic drug development, he recommends a combination of "push" and "pull" interventions by governments.

Push subsidies would help fund early-stage drug development, getting more antimicrobials into the drug approval pipeline. Pull funding would guarantee large payments or purchases to companies that bring new drugs to market. To cover the cost of the new interventions, Hall suggests one option would be charging a fee to pharmaceutical companies based on whether they invest in antibiotic development; those that are putting sufficient money into new antimicrobials might be exempt from such a fee, with the revenues used to finance the new subsidies.

New antibiotics will suffer the same fates as older ones unless companies and policymakers also work harder on antibiotic stewardship. It's often cheaper and easier for physicians to simply prescribe antibiotics without checking whether a patient's infection is susceptible to them. That increases the risk of promoting resistance. To break this pattern, Hall's team advocates mandatory diagnostic testing, in high income countries that can afford it, to reduce inappropriate antibiotic use.

Returning to the meeting's main topic, Hall concedes that the evidence for the impact of agricultural antimicrobial use remains unsettled. "There's obviously going to be a debate for many years to come about the extent of the threat to human health from antibiotic use in agriculture, but we feel that there is enough evidence now to start making steps," he said. In particular, he advocates setting targets for reducing the use of antimicrobials in agriculture, agreed in principle internationally, but set transparently by individual governments for thier own country. This would allow countries to develop their own policies to meet those targets.

Data gap

After Hall's presentation, all of the speakers from the second session returned to the stage for an extended question session. The discussion began with a question about rising consumer demand for "antibiotic-free" foods. Sneeringer explained that it is still hard to pinpoint how much more people are willing to pay for such claims, though her group is now collaborating with supermarkets to collect data on that.

The lack of data on other aspects of antimicrobial resistance clearly frustrated other speakers and many members of the audience, and the group talked about the difficulties of reaching international agreements on drug use when many nations don't even track their current usage rates. Nonetheless, Hall expressed optimism that if the global community can agree on reduction targets, many nations could implement monitoring programs relatively easily.

An even bigger mystery is what effect the huge quantities of antibiotics in agricultural waste have on the soil microbiomes around farms. Fanning pointed to one recent study that did a preliminary analysis of the changes, but added that, "we don't understand what causes that adaptive mechanism within bacteria to cause some [antibiotic resistance] to be really problematic and others that are more transient."

All three speakers discussed the challenge of keeping industry marketing claims grounded in reality, especially when many prominent commentators in the food business promote dubious ideas. Marsden recommended that scientific organizations take a more active role in public discussions of food safety and nutrition.

Despite their diverse backgrounds, speakers and members of the audience largely agreed that antimicrobial stewardship in agriculture is at a crossroads. As Hall said in his presentation, "there's a lot of momentum at the moment, but if we don't have some sort of [regulatory] agreements this year it really feels like we might lose that momentum."