Antimicrobial Resistance: Implications for the Food System Microorganisms have an inherent ability to evolve—to mutate and adapt to environmental stressors─allowing them to survive otherwise lethal conditions.

IFT convened a panel of internationally renowned experts to address the concern that the use of antimicrobials in food production, manufacturing, and elsewhere may lead to the emergence of foodborne pathogens that are resistant to antimicrobials, thus compromising the ability to subsequently control them, whether in production agriculture, food processing, or human medicine.

August 2, 2006


The safety of food worldwide remains challenged by the potential for emergence of new pathogens and re-emergence of known pathogens. Microorganisms have an inherent ability to evolve–to mutate and adapt to environmental stressors–allowing them to survive otherwise lethal conditions. The Institute of Food Technologists (IFT),1 the 22000-member nonprofit scientific and educational society, convened a panel of internationally renowned experts to address the concern that the use of antimicrobials in food production, manufacturing, and elsewhere may lead to the emergence of foodborne pathogens that are resistant to antimicrobials, thus compromising the ability to subsequently control them, whether in production agriculture, food processing, or human medicine. The outcome of the panel's deliberations is presented in this Expert Report. IFT's objective for this Expert Report is to increase the understanding–among IFT members, senior policy officials, and other interested groups–of the state of the science on the public health impact of the use of antimicrobials in the food system, and development and control of antimicrobial resistance. This report is the fourth Expert Report produced by IFT.

DIGITAL OBJECT IDENTIFIER (DOI) 10.1111/j.1541-4337.2006.00004.x


The availability of antibiotics to treat infectious diseases has radically improved human and animal well being, and to a minor degree, plant health. Paradoxically, this very success threatens the future utility of antibiotics. The discovery of penicillin in 1940 ushered in the era of "modern medicine." Numerous antimicrobials, including most structural classes of antibiotics were discovered during 1920 to 1970. Chemical modification of many of these compounds led to new entities with superior activities. Because of the great success in antibiotic discovery, by the late 1970s, many proclaimed that the war on infectious diseases had been won, leading ultimately to de-emphasis of antibiotic discovery during the 1980s and a decline in the 1990s. At the same time, however, widespread antibiotic resistance was emerging and resulting in impaired treatment of human diseases (Neu 1992). As the genomes of bacteria, especially pathogens, have become increasingly available, the prospect of using them to identify new targets for antibiotic discovery has renewed interest in such investigations between the public sector and large pharmaceutical and biotechnology companies. Many of the larger companies and much of the public sector, however, have redirected research efforts toward noninfectious disease targets.

All uses of antibiotics in human medicine and animal husbandry create selective pressure that favors emergence of antibiotic resistance among microorganisms, which could undermine the effectiveness of the antibiotics and potentially give rise to a "postantibiotic" era. The selection for antibiotic-resistant bacteria in agricultural production environments and the subsequent impact on animal and human health has become a major concern and is the subject of many reports (Table 1). This document focuses on the use of antimicrobial agents to control bacteria in the food system; other microorganisms are considered as well, however. This document builds upon the IFT Scientific Status Summary "Resistance and Adaptation to Food Antimicrobials, Sanitizers, and Other Process Controls" (IFT 2002a), to inform readers about the various types of antimicrobial agents, including antibiotics, food antimicrobial agents, and sanitizers that are used at various stages of the food system, and the mechanisms that microorganisms, particularly foodborne pathogens, have for surviving the stress of exposure to these substances in their environments. Trends in antimicrobial resistance, and the resultant human health, economic, and clinically relevant environmental impacts are also addressed.

Table 1–Reports of antimicrobial use, resistance, and human health impact

DateCountry or InternationalReport sourceReport titleURL address (if applicable)
1969 United Kingdom English Parliament The Report to Parliament by the Joint Committee on Antibiotic Uses in Animal Husbandry and Veterinary Medicine ("Swann Report")
1980 United States National Research Council (NRC) The Effects on Human Health of Subtherapeutic Use of Antimicrobials in Animal Feed
1981 United States Council for Agricultural Science & Technology Antibiotics in Animal Feeds, Report 88
1981 United States Institute of Medicine (IOM) Human Health Risks with the Subtherapeutic Use of Penicillin or Tetracyclines in Animal Feed
1989 United States IOM Committee on Human Health Risk Assessment of Using Subtherapeutic Antibiotics in Animals Human Health Risks with the Subtherapeutic Use of Penicillin or Tetracyclines in Animal Feeds
1997 International World Health Organization (WHO) The Medical Impact of the Use of Antimicrobials in Food Animals
1998 United Kingdom UK Ministry of Agriculture, Fisheries, and Food A Review of Antimicrobial Resistance in the Food Chain
1998 United States United States Food and Drug Administration (FDA) Center for Veterinary Medicine (CVM) A proposed framework for evaluating and assuring the human safety of the microbial effects of antimicrobial new drugs intended for use in food-producing animals
1998 International WHO Use of Quinolones in Food Animals and Potential Impact on Human Health: Report and Proceedings of a WHO Meeting
1999 European Union The European Agency for the Evaluation of Medicinal products Antibiotic Resistance in the European Union Associated with Therapeutic Use of Veterinary Medicines
1999 European Union EU Scientific Steering Committee Opinion of the Scientific Steering Committee on Antimicrobial Resistance

1999 United States FDA Risk Assessment on the Human Health Impact of Fluoroquinolone-resistant Campylobacter Associated with Consumption of Chicken

(revised as of January 5, 2001)
1999 United States NRC National Academy of Sciences Committee on Drug Use in Food Animals and the Panel on Animal Health, Food Safety, and Public Health The Use of Drugs in Food Animals: Benefits and Risks
1999 United States U.S. General Accounting Office (GAO) Food Safety: The Agricultural Use of Antibiotics and Its Implications for Human Health
1999 United Kingdom Advisory Committee on the Microbiological Safety of Food Report on Microbial Antibiotic Resistance in Relation to Food Safety
(a synopsis)
1999 Australia Joint Expert Advisory Committee on Antibiotic Resistance The Use of Antibiotics in Food-Producing Animals: Antibiotic Resistant Bacteria in Animals and Humans
1999 European Union European Commission Opinion of the Scientific Steering Committee on Antimicrobial Resistance, May 28, 1999
1999 International WHO The Medical Impact of the Use of Antimicrobials in Food Animals
2000 United States Centers for Disease Control and Prevention Interagency Task Force on Antimicrobial Resistance A Public Action Health Plan to Combat Antimicrobial Resistance
2000 International WHO WHO Global Principles for the Containment of Antimicrobial Resistance in Animals Intended for Food
2000 International Food and Agriculture Organization of the United Nations (FAO)/WHO Codex Committee on Residues of Veterinary Drugs in Foods Antimicrobial Resistance and the Use of Antimicrobials in Animal Production
2001 International Office International Des Epizooties (OIE) Antimicrobial Resistance: Reports prepared by the OIE Ad Hoc Group of Experts on Antimicrobial Resistance
2001 International WHO WHO Global Strategy for Containment of Antimicrobial Resistance
2001 International WHO Monitoring Antimicrobial Usage in Food Animals for the Protection of Human Health
GSSMontitoring AMRuseOslo.pdf
2002 United States Alliance for the Prudent Use of Antibiotics The Need to Improve Antimicrobial Use in Agriculture: Ecological and Human Health Consequences ("FAAIR Report")
2002 Canada Veterinary Drugs Directorate, Health Canada Report of the Advisory Committee on Animal Uses of Antimicrobials and Impact on Resistance and Human Health Uses of Antimicrobials in Food Animals in Canada: Impact on Resistance and Human Health
2003 International WHO Department of Communicable Diseases, Prevention and Eradication and Collaborating Centre for Antimicrobial Resistance in Foodborne Pathogens Impacts of Antimicrobial Growth Promoter Termination in Denmark
2004 International FAO, OIE, and WHO Joint FAO/OIE/WHO Workshop on Non-Human Antimicrobial Usage and Antimicrobial Resistance: Scientific Assessment
2004 International FAO, OIE, and WHO Second Joint FAO/OIE/WHO Expert Workshop on Non-Human Antimicrobial Usage and Antimicrobial Resistance: Management Options
2004 United States GAO Federal Agencies Need to Better Focus Efforts to Address Risk to Humans from Antibiotic Use in Animals

Classification of antimicrobials

For the purposes of this report, "antimicrobial" is a general term used broadly to refer to any compound, including antibiotics, food antimicrobial agents, sanitizers, disinfectants, and other substances, that acts against microorganisms. The definitions and use of each of these terms differ among various groups. Legal definitions exist for use in a regulatory context.

The term antibiotic is used in this report to refer to drugs used to treat infectious disease in humans, animals, or plants, by inhibiting the growth of or destroying microorganisms; such substances may be naturally occurring, semisynthetic, or synthetic. Antibiotics are also used in food animals to prevent infectious disease and improve the efficiency of feed utilization. Within the antibiotic classification are synthetic antimicrobials such as quinolones, that differ from other substances such as streptomycin, which are natural products or fermentation derived antibiotics. Antibiotics are legally classified as such only when used in humans. They are classified as "veterinary antimicrobial drugs" when used in animals and as "pesticides" when used in plants.

"Biocide" is a general term that refers to chemical agents, such as disinfectants and sanitizers, which are usually broad spectrum. Because biocides vary in antimicrobial activity, other terms may be used to more specifically describe the nature of the antimicrobial activity. For example, terms ending in the suffix "-static," such as "bacteriostatic," are used for agents that inhibit microbial growth without killing the microbes, and terms with the suffix "-cidal," such as "fungicidal," refer to agents that kill the target microbe (McDonnell and Russell 1999). "Disinfectants" destroy or irreversibly inactivate infectious fungi and vegetative bacteria (growing or nonsporeforming), and are used in hospitals, food processing facilities, restaurants, and elsewhere for general purposes (EPA 2005). In the legal connotation, disinfectants include "any oxidant, including but not limited to chlorine, chlorine dioxide, chloramines, and ozone, added to water in any part of the treatment or distribution process, that is intended to kill or inactivate pathogenic microorganisms" (40 CFR §141.2). "Sanitizers," comprised of 2 categories–no-rinse food contact surface sanitizers and nonfood contact surface sanitizers–refer to substances that reduce microbial contamination and destroy vegetative pathogens of public health significance on treated inanimate surfaces. "Sterilants," such as peroxyacetic acid, refer to substances that eliminate all forms of microbial life, including bacterial spores, fungi, and viruses. IFT uses the legal connotation "food antimicrobial agent" to refer to antimicrobial substances, such as nisin and other bacteriocins, including mold inhibitors, which are used to preserve food by preventing microbial growth and subsequent spoilage. Antibiotics cannot legally be used as food additives; thus, they are specifically excluded from this classification.

Classification of resistance

Discussion of antimicrobial resistance, by necessity, must include defining what is meant by resistance. While it would seem that defining resistance would be a simple matter, many definitions exist (Davison and others 2000). Resistance to most traditional, regulatory-approved, or naturally occurring food antimicrobial agents is difficult to characterize because of the lack of a precise definition for such resistance. From a functional perspective, resistance correlates with failure of a given antimicrobial treatment, whereas from a laboratory perspective, resistance is denoted through a "Minimal Inhibitory Concentration" (MIC)2 value that exceeds a threshold value, which may or may not be associated with a clinical outcome. Chapman (1998) stated that a microorganism is resistant if it exhibits "significantly reduced susceptibility" when compared with that of the "original isolate" or a group of sensitive strains. In this report, resistance means "temporary or permanent ability of a microorganism and its progeny to remain viable and/or multiply under conditions that would destroy or inhibit other members of the strain" (Cloete 2003). These terms and the different types of resistance are described below.

As Courvalin (2005) describes, resistance can result from mutations in housekeeping structural or regulatory genes, or alternatively, horizontal acquisition of foreign genetic information. In some cases, resistance may manifest through multiple mechanisms. For example, three different strategies are thought to be involved in resistance to tetracycline (Schnappinger and Hillen 1996). Resistance can also be intrinsic, that is microorganisms without known exposure to antimicrobial agents may be resistant to some agents (see below).

If a resistant strain is isolated from an environment containing an antimicrobial or prepared in the laboratory by exposure to increasing concentrations of an antimicrobial, resistance may be due to a genetic alteration or a temporary adaptation. It may be that temporary adaptation to an antimicrobial through some type of homeostatic mechanism plays a much larger role than true genetic mutation among food-related antimicrobials. To date, research on resistance to food antimicrobials has focused almost exclusively on innate or intrinsic mechanisms of the target microorganisms.

Innate (Intrinsic) resistance. As is the case for a natural property of a microorganism, innate resistance is chromosomally controlled (Russell 1991). Innate resistance is related to the general physiology or anatomy of a microorganism and stems from pre-existing mechanisms or properties. This type of resistance is most likely responsible for differences in resistance observed among different types, genera, species, and strains of microorganisms in identical environmental conditions and concentrations. Innate resistance may stem from the complexity of the cell wall, efflux mechanisms (means by which microbes pump antimicrobials out of the cell [Gilbert and McBain 2003]), or enzymatic inactivation of the antimicrobial (Russell 2001). For example, because of the complexity of their cell walls, Gram-negative bacteria generally have a higher level of resistance to antibacterial agents than typical Gram-positive bacteria (Russell and Chopra 1996). More specifically, Gram-negative bacteria are innately resistant to penicillin G by virtue of their double membrane structure, which prevents the antibiotic from accessing the cell wall target. Similarly, Mycobacterium species are more resistant than other nonsporeforming bacteria due to high lipid content in their cell walls and comparatively high hydrophobicity. Other bacteria–Bacillus, Pseudomonas, Corynebacterium, Micrococcus, and the fungus Aspergillus have innate resistance to benzoate because they are capable of metabolizing the compound to succinic acid and acetyl coenzyme A (Chipley 1993). Innate resistance is not considered an important clinical problem because antibiotics were never intended for use against intrinsically resistant bacteria.

There are certain circumstances in which antimicrobial agents do not adversely affect bacteria that are generally susceptible to the particular agent. Because the efficacy of most food antimicrobials and sanitizers is dependent upon and influenced by the conditions of the application, some situations may permit bacterial resistance that would not have occurred otherwise (IFT 2002a). Exposure conditions, such as the environmental conditions (temperature, pH, and food composition) of the antimicrobial application, or interaction of the antimicrobial with components of the suspension medium or food product can influence the efficacy of the antimicrobial agent (Davidson 2001). For example, organic matter can quench the hypochlorite ion and therefore eliminate its efficacy at killing generally susceptible bacterial populations (Kotula and others 1997). However, microorganisms that are generally susceptible to antibiotics can themselves also become temporarily resistant to an antimicrobial through activation of silent, resident gene(s) that confer this resistance. A good example of this occurring is observed with the survivability of biofilm-associated cells versus planktonic (free floating/living) cells (Frank and Koffi 1990; Mosteller and Bishop 1993; Mustapha and Liewen 1989). Microbial cells in biofilms exhibit resistance primarily through the protection provided by extracellular materials such as exopolysaccharides. Also, nongrowing bacterial cells are resistant to many antibiotics that target cell wall synthesis. Once conditions again become favorable for growth, these bacterial cells become susceptible again to these cell wall inhibitors. Also, the few reports of resistance to food antimicrobial preservatives and sanitizers are attributed to microbial stress responses to sublethal stressors, such as low or high temperatures, acidity, osmolarity, low moisture, high atmospheric pressure, low oxygen or anaerobic conditions, gas atmospheres, competing bacteria, and low nutrient environments that trigger physiological changes and subsequently confer resistance to these compounds (Abee and Wouters 1999; Archer 1996; Samelis and Sofos 2003a; Sheridan and McDowell 1998; Sofos 2002a).

Acquired (Extrinsic) resistance. Acquired resistance results from genetic changes that occur through mutation of the antimicrobial's target site within the bacterium or acquisition of genetic material encoding resistance via plasmids3 or transposons4 containing integron sequences5 (Roe and Pillai 2003; Russell 1991, 1996; Russell and Chopra 1996). Acquired resistance, the most common type of antibiotic resistance, has been well studied for antibiotics, but has not been well studied for food antimicrobial agents and sanitizers. Acquisition of genes for β-lactamase (an enzyme capable of breaking down and inactivating β-lactam antibiotics [penicillins and cephalosporins]) and mutation of one of the subunits of DNA gyrase (the target of fluoroquinolones) are examples of this type of resistance. Another example includes resistance of some microorganisms to sanitizing compounds, such as quaternary ammonium compounds (QACs), as a result of the presence of plasmid-encoded efflux pumps that remove the QACs (Russell 1997).

Although acquired resistance is of concern in the use of food antimicrobial agents and sanitizers, occurrence of such resistance appears to be rare. Unlike antibiotics, which generally have specific target sites, biocides (that is, disinfectants, sanitizers, antimicrobials) typically act nondiscriminately against multiple nonspecific targets (Bower and Daeschel 1999); thus, single mutations or biochemical alterations of cellular targets seldom confer resistance to biocides.

Adaptation. For certain types of antimicrobials, adaptation may be demonstrated by exposing a microorganism to a stepwise increase in concentration of the substance. This type of resistance, however, is often unstable; the microorganism may revert back to the sensitive phenotype when grown in an antimicrobial-free medium, termed "back-mutation" (Russell 1991). In the absence of selection pressure, the mutations associated with resistance may actually reduce fitness of the bacterial strain compared to the wild type, parental strain. Stabilizing, secondary, compensatory mutations are sometimes needed to maintain resistance and reduce fitness cost associated with the original "resistance" mutation.

Antimicrobial Applications

During food production and manufacturing, a variety of antimicrobials, including antibiotics, antifungals, sanitizers, and food preservatives, are applied to improve the efficiency of the system, and increase the safety and quality of the product. The multiple points throughout the ecosystem where antimicrobials may be used and subsequently impact the epidemiology of resistance are shown in Figures 1 and 2. Microorganisms encounter and are subjected to a variety of antimicrobial stressors as they move throughout the food system, from the environment to the plant, through food processing, shipping, distribution, storage, and into kitchen food preparation areas. The variety of antimicrobial uses at each of the various stages of the food system may create selective pressure that promotes resistance.

Epidemiology of antimicrobial resistance (adapted and modified from Linton [1977] by Rebecca Irwin, Health Canada [Prescott 2000] and IFT, with permission).
Figure 1–Epidemiology of antimicrobial resistance (adapted and modified from Linton [1977] by Rebecca Irwin, Health Canada [Prescott 2000] and IFT, with permission).
Application of antimicrobials from farm to table
Figure 2–Application of antimicrobials from farm to table

The major classes of antibiotics and their various uses in animals, plants, and humans are listed in Table 2. Detailed information on the mechanism of action of specific classes of antimicrobials can be found elsewhere (Prescott and others 2000; Walsh and others 2003). Some of the antibiotics and fungicides used in agriculture have identical chemical counterparts in human medicine. The majority of antibiotics used in food animals belong to classes of antibiotics that are also used to treat human illness; these include tetracyclines, sulfonamides, penicillins, macrolides, fluoroquinolones, cephalosporins, aminoglycosides (gentamicin and kanamycin), chloramphenicols, and streptogramins (NARMS 2006).

Table 2–Examples of antimicrobial drugs and antibiotics, by major class, approved in the United States for animal, plant, or human use

Antimicrobial, drug class (selected examples)Mode of action/spectrumFood animal usePlant useHuman use

Animal SpeciesDisease treatmentDisease preventionGrowth promotion
Aminoglycosides (gentamycin, neomycin, streptomycin) Inhibit protein synthesis/broad spectrum Beef cattle, goats, poultry, sheep, swine
• (Certain plants)
Beta-lactams penicillins (amoxicillin, ampicillin) Inhibit cell wall synthesis Beef cattle, dairy cows, fowl, poultry sheep, swine
cephalosporins 1st generation (cefadroxil) Broad spectrum
cephalosporins 2nd generation (cefuroxime)

cephalosporins 3rd generation (ceftiofur) Beef cattle, dairy cows, poultry, sheep swine

Chloramphenicol Inhibit protein synthesis/
(Florfenicol) broad spectrum

Inhibit protein synthesis/broad spectrum Beef cattle
Cycloserines (cycloserine) Inhibit cell wall synthesis/narrow spectrum
Glycopeptides (vancomycin) Inhibit cell wall synthesis/narrow spectrum
Ionophores (monensin, salinomycin, semduramicin, lasalocid) Disrupts osmotic balance/narrow spectrum Beef cattle, fowl, goats, poultry, rabbits, sheep
Lincosamides (lincomycin) Inhibit protein synthesis/narrow spectrum Poultry, swine
Macrolides (tylosin, tilmicosin erythromycin) Inhibit protein synthesis/narrow spectrum Beef cattle, poultry, swine
Monobactrams (aztreonam) Inhibit cell wall synthesis broad spectrum
Polypeptides Inhibit cell Fowl, poultry,
(bacitracin) wall synthesis narrow spectrum swine

Fluoroquinolones Inhibit DNA Beef cattle
(enrofloxacin, danofloxacin) synthesis/broad spectrum
Streptogramins (virginiamycin) Inhibit protein synthesis/narrow spectrum Beef cattle, poultry, swine
Sulfonamides (sulfadimethoxine sulfamethazine sulfisoxazole) Inhibit folic acid synthesis/broad spectrum Beef cattle, dairy cows, fowl, poultry, swine, catfish, trout, salmon

Tetracyclines (chlortetracycline oxytetracycline tetracycline) Inhibit protein synthesis/broad spectrum Beef cattle, dairy cows, fowl, honey bees, poultry sheep, swine, catfish, trout, salmon, lobster • (Certain plants)
Bambermycin Inhibit cell wall synthesis/narrow spectrum Beef cattle, poultry, swine

Carbadox Inhibits DNA swine synthesis/narrow spectrum

Novobiocin Inhibits DNA gyrase/narrow spectrum Fowl, poultry
Spectinomycin inhibit protein synthesis/narrow Poultry, swine

aPoultry includes at least one of the following birds: broiler chickens, laying hens, and turkeys.
bFowl includes at least one of the following birds: ducks, pheasants, and quail.
(adapted and modified from GAO 1999)

Antibiotics are also used in companion animals, most often for treating dermatological conditions, ear infections, respiratory infections, urinary tract infections, and wounds. Applications in companion animals are addressed in Appendix 1.

Production agriculture

Animal husbandry. Foods of animal origin have been a mainstay of American agriculture. During the past half-century, food animal production has increased dramatically as a result of advances in science and technology, including the use of antibiotics in treating and preventing disease. Improvements in animal genetics, housing, nutrition, biosecurity, husbandry, and veterinary medicine, concurrent with more efficient business practices and economies of scale, have allowed food animal production to meet the demands of consumers. Antibiotics have been used in food animals (primarily cattle, swine, and poultry) for more than 50 years to treat, prevent, or control infectious disease, or to improve efficiency of feed utilization and weight gain (Gustafson and Bowen 1997). Specific information on antimicrobial agents used in animals can be found in the "Green Book" (listing the FDA-approved animal drug products) or the Feed Additive Compendium (Anonymous 2006a; FDA/CVM 1998). Administration of these veterinary drugs to food animals is a critical component of an overall management system that food animal producers use to secure the health and welfare of the animals and ensure the safety of the products that enter the food chain. Commensurate with increased food animal productivity is the inevitable shift to more intensive production systems, most notably in beef cattle, poultry and swine, to meet the expectations and needs of a growing number of people. Antibiotic use in food animals as an overall strategy to prevent and treat infectious disease is most relevant to antibiotic use in intensive production systems, in which the health of food animals is linked to consumer need for plentiful amounts of food animal products, food safety, and public health.

In modern production systems, food animals are generally raised in groups (NRC 1999). Typically chickens are raised in barns accommodating 10000 to 20000 birds, pigs are maintained in multiple-pen buildings, and beef cattle are raised outdoors in large pens in feed yards. Given the close proximity of the animals to one another (commingling), physiological and environmental stressors, and immature immune systems, any underlying viral infections, or bacterial respiratory or enteric diseases that may occur in a few animals can spread to others including entire herds of flocks. Within the limits of the production system, and depending on the nature of the disease, the producer and/or veterinarian may intervene in such situations by medicating the entire group via the feed or water rather than treating each affected animal. Feed medication is more efficient for long-term prophylaxis, whereas medication of water is more effective for treating disease outbreaks due to its rapid intake and clinical response elicitation. Medicated water is also a more effective means for treating sick animals, which often continue to drink despite not continuing to eat. Administration of medication via water also allows large numbers of animals to be treated in an efficient manner, and avoids worker safety issues associated with injecting large numbers of animals.

Therapeutic uses. Therapeutic antimicrobial regimens include treatment, control, and prevention of disease (NCCLS 2002). Treatment is the administration of an antimicrobial to an animal or group of animals exhibiting frank, clinical disease (NCCLS 2002). Control is the administration of an antimicrobial to animals, usually as a herd or flock (metaphylaxis), in which morbidity and/or mortality has exceeded baseline norms, that is, early in the course of disease onset in the population. For example, as beef calves arrive at the feedlot, some of the animals disembarking from the truck may exhibit signs of clinical disease, for which treatment is necessary. While the other animals from the truck appear healthy, they have likely been exposed to the inciting pathogen and would otherwise "break" with disease if not also treated. The control concept is based on the premise that because the risk of disease spread from an individual animal or small group of diseased animals to the large susceptible population is substantial, it is appropriate that all animals be medicated. Prevention or prophylaxis is the administration of an antimicrobial to exposed at-risk healthy animals, generally in a herd or flock situation rather than on an individual animal basis, prior to the onset of a disease for which no etiologic agent has been cultured. An example of antimicrobial prophylaxis is the intramammary infusion of antibiotics to all dairy cows in a herd at the end of the lactation cycle, known as "dry-cow therapy," to prevent mastitis at parturition.

Occurrence of risk factors for a particular disease, herd/flock history, and the appearance of clinical signs in some animals may be sufficient indication that empirical antibiotic therapy is warranted to limit potential spread among an animal population. Empirical treatment is based upon the experience of the veterinarian or food animal producer, and involves consideration of such factors as animal species and its susceptibility to suspected pathogen(s), pathogen virulence, treatment cost, and any applicable antibiotic withdrawal times.6 In such circumstances, a bacteriological diagnosis is most often made retrospectively from a necropsy specimen from a dead animal, although a culture from a live animal within the exposed population sometimes is recognized by analysis. Upon pathogen identification, the diagnostic laboratory will perform antimicrobial susceptibility testing, the results of which will further guide the veterinarian in antibiotic selection.

All of the newer injectable and water-soluble antibiotic products, including ceftiofur, enrofloxacin, and florfenicol, must be obtained by prescription from or dispensed by a veterinarian. Antibiotic agents intended for growth promotion or therapeutic use in feed are usually incorporated into the feed at the feed mill and fed directly to the animals without direct veterinary involvement. An exception, however, is a prescription-like order signed by a veterinarian, through the Veterinary Feed Directive, that is processed and "filled" at the feed mill.

Extra-label7 drug use is also a legal option for specific circumstances in food animal production. Such use of a drug differs from its approved labeling, which addresses species, indication, dosage levels, and frequency or route of administration. Under strict provisions that include a veterinarian-client-patient relationship, veterinary uses of extra-label drugs are acceptable to the FDA as long as the regulatory requirements are met, including that any tissue residues of the drug in meat or meat products are less than predetermined limits. Extra-label drug use, however, is not permitted for drugs added to feed (21 CFR §530).

Performance improvement uses. In the 1950s, it was shown that antibiotics administered at low levels for an extended time period promote growth rate and feed efficiency (growth promotion) in healthy livestock, primarily cattle, swine, and chickens (Jukes 1971). The beneficial effects of antibiotics on feed efficiency and growth rate have since been demonstrated for all major livestock species (Hays 1991). The use of an in-feed antibiotic for growth promotion occurs most often in young, growing animals. Use in older animals has a lessened effect. The use of antibiotics for growth promotion is intended to allow farmers to produce food animals at less cost because the amount of feed required for an animal to reach production weight is reduced.

A number of mechanisms for the growth promotion effects of antibiotics have been proposed. Possible mechanisms have been reviewed by Gaskins and others (2002) and Shryock (2000). The potential mechanisms are thought to be physiological, nutritional, and metabolic in nature and relate to antibiotic inhibition of the normal microflora, enabling more energy to be expanded for nutrient use and increased conversion to weight gain. Studies with germ-free animals have suggested that growth promotion results from antibacterial activities within the gastrointestinal system (Feighner and Dashkevicz 1987). Since the only known common factor among the various structurally and mechanistically distinct antibiotics used for growth promotion is the ability to kill bacteria, this mechanism seems plausible. Further, the three antibiotics (tetracycline, tylosin, and bacitracin) most commonly used for growth promotion act by inhibiting bacterial protein or cell wall synthesis. Moreover, the intestinal microflora of animals affects gut physiology in a number of ways, influencing for example, water uptake, immune response, and nutrient availability (Savage 1977).

Collier and others (2003) found that tylosin decreased total bacteria within the digestive tract and reported that the decrease may reduce host-related intestinal or immune responses, which would divert energy that could otherwise be used for growth. Modulation of the intestinal microflora of animals, resulting in selective enrichment for certain "optimal" bacteria, could enhance gut physiology by optimizing metabolism or nutrient uptake. Thus, it is also thought that the optimal microflora assist in maintaining animal health, and subsequently public health as well, by selectively excluding pathogens through either occupation of the physical intestinal microhabitat or acting as microbial antagonists. Collier and others (2003) also reported that the ability of tylosin to improve animal growth may relate to its apparent selection for lactobacilli, commensals8 known to competitively exclude potentially pathogenic species from colonizing the intestine.

The use of antibiotics for growth promotion, however, has been a target for elimination. In the European Union (EU), growth promotion claims for human use class feed additive antibiotic labels were withdrawn in the 1990s, and nonhuman use class feed additive antibiotics followed in January 2006. In the United States some large restaurant corporations (for example, McDonalds, Oak Brook, Ill., U.S.A.) have developed antibiotic use policies that exclude human-use antibiotic classes for growth promotion purposes in flocks and herds of suppliers from whom they purchase poultry and beef products.

Poultry. The poultry industry is the most integrated of all of the major food animal industries in the United States. With integration, a single company controls the entire production cycle, from breeders to retail market. Approximately 8.4 billion chickens (broilers) and 264 million turkeys were produced in 2004 (USDA/NASS 2005). In most hatcheries, day-old chicks are injected with vaccines or an antibiotic, such as gentamicin or ceftiofur, to prevent opportunistic bacterial infections. Broiler chickens (typically six to eight weeks of age and five to eight pounds) are typically raised in pens containing 10000 to 20000 birds; turkeys are typically raised in groups of 5000 to 10000 (Lasley 1983; Lasley and others 1983). The majority of drugs used in poultry are administered via feed or water. Ionophores9 or arsenicals are used as coccidiostats and antibiotics are used as growth promoters (NRC 1999).

Starter and grower rations may contain up to three drugs–a prophylactic coccidiostat, an antibiotic growth promoter, and an arsenical compound having both anticoccidiostat and growth-promoting properties. One or more drugs may be deleted from grower and finisher rations, however, to reduce cost and comply with drug withdrawal times to prevent tissue residues (NRC 1999). Table 2, which lists the antimicrobials approved for use in the United States, identifies a number of antibiotics (for example, bacitracin, bambermycin, chlortetracycline, penicillin, and virginiamycin) that are approved for use for growth promotion and feed efficiency in broilers, turkeys, and layers. Several antibiotics, administered as feed additives, are approved for treating intestinal infections, such as necrotic enteritis (caused by Clostridium perfringens) and coccidiosis (a common parasitic poultry disease caused by Eimeria species). Bacitracin and virginiamycin, for example, are used to treat necrotic enteritis, and monensin, salinomycin, narasin, and semduramicin are used to treat coccidiosis. Respiratory disease, such as air sacculitis caused by Escherichia coli, is treated with tetracycline. A variety of other antimicrobial agents are used for various conditions in poultry production (Merck 2003).

Swine. In 2004, 103 million hogs were slaughtered for food use (USDA 2005a). To control their environment and reduce disease, swine are often raised in confinement, sometimes from birth to slaughter (farrow to finish), or in age-segregated management systems where they are moved to different farms for various production stages (nursery, grower, and finishing, for example). Increasingly, management systems are undergoing transition to the all-in/all-out system in which pigs of similar ages are housed together to limit spread of infectious disease among animals with different age-dependent immune systems. Operations with more than 5000 head accounted for more than 75% of the swine in the United States in 2001, compared with only 27% in 1994 (USDA 2003). Several major pork production companies are fully integrated, but most production is still segmented.

The majority of drugs for swine are administered via feed or water. Breeding sows and pre-weaning pigs, however, are an exception, with antibiotics generally administered to individual animals. Most swine receive an antibiotic in feed ("starter rations") after weaning, when they are most vulnerable to infectious disease (caused by antecedent viral infections predisposing the animals to mycoplasma and/or bacterial superinfection) that may be related to the stress of weaning and movement within the production unit (Dewey and others 1997). Pneumonia is an important problem in swine production; antibiotics such as ceftiofur, tilmicosin, penicillin, lincomycin, tetracyclines, and tiamulin are used to treat and prevent clinical cases and outbreaks. Gentamicin, carbadox, tetracyclines and neomycin are sometimes used to control diarrhea caused by bacteria such as E. coli and C. perfringens. Ileitis (caused by Lawsonia intracellularis) may be treated with antibiotics such as lincomycin, tiamulin, or tylosin. Feed efficiency and growth promotion can be achieved with bacitracin, tylosin, virginiamycin, tetracyclines, and penicillin. A variety of other conditions, for which other antimicrobial agents are used, exist in swine production (Merck 2003).

Beef cattle. More than 37 million head of cattle were slaughtered in 2004 (USDA 2005a). In contrast to the highly integrated poultry industry, the beef cattle industry is still quite segmented, with many calves changing ownership and shipped multiple times during their lifetime. Calves from many sources are combined via auction or sale barns, transported, and commingled at the feed yard. Upon entering a feedlot, young cattle are given vaccinations against gastrointestinal and respiratory diseases, as well antihelminthic drugs. During stressful events, such as weaning or transportation and commingling, calves often develop pneumonia or diarrhea–major causes of mortality–and are often treated via individual or group medication.

In the U.S. beef industry, the majority of antibiotics are used on feedlots (USDA 2000). In 1999, the U.S. Dept. of Agriculture (USDA) conducted a survey of U.S. feedlots to determine antibiotic treatment practices. For treatment of individual animals, approximately 50% of feedlots used tilmicosin and/or florfenicol and/or tetracyclines as part of the initial therapy. The feedlots also used cephalosporins (38.1%), penicillins (31.1%), fluoroquinolones (32.1%), and macrolides (17.4%) for individual animal therapy. Approximately 41% of feedlots administered antibiotics for metaphylactic therapy; those most commonly used were tilmicosin, oxytetracyclines, and florfenicol (among 70.3%, 31.9%, and 22.1% of feedlots, respectively; USDA 2000). An estimated 83% of feedlots administered at least one antibiotic to cattle in feed or water for disease prophylaxis (tylosin for liver abscesses, for example) or to increase feed efficiency. A variety of other antimicrobial agents are used for a variety of conditions in beef cattle production (Merck 2003).

Dairy cattle and veal calves. There were 9.12 million cattle in dairy production in 2001 (USDA 2002). Dairy herd health is closely associated with milk production and economic sustainability. Therefore, maintenance of herd health is closely dependent upon disease prevention and therapeutic drug use for a range of diseases. Severe diarrhea and pneumonia are two main causes of morbidity and mortality in dairy heifers. Most dairy heifers are vaccinated against a range of gastrointestinal and respiratory diseases to minimize the need for antibiotics. Other conditions such as footrot and reproductive diseases may require antibiotic treatment specific to the diagnosis (Merck 2003). Administration of antibiotics to lactating cows, however, must be done with care to avoid milk residues. Mastitis is the most costly disease among dairy cattle, and intramammary infection is the most costly disease in U.S. food animal production (NRC 1999). Acute mastitis must be diagnosed in individual cows and can be treated with intramammary infusions of several antibiotics, for example, β-lactams, pirlimycin, and erythromycin. Except for mastitis caused by environmental pathogens (coliforms, for example), which does not always require antibiotic therapy, antibiotics to prevent mastitis are often administered through intramammary infusions at the beginning of the "dry (nonlactating) period" on a routine basis to all animals in the herd (Gibbons-Burgener and others 2000).

To reduce transmission of disease from the dam, the majority of dairy calves are separated from dams within 24 hours of birth and provided an initial feeding of colostrums, often pasteurized, from the initial milking to provide maternal antibodies and immunity. Most calves are housed in individual hutches or pens to control infection, and are fed milk or milk replacers (that may be medicated with an antibiotic) until weaning at 6 to 8 weeks of age, after which time they are generally housed in groups. The males and excess females are sometimes used for veal production.

The majority of veal calves are raised in the United States individually in stalls until they are 16 to 18 weeks of age. Due to their young age and confinement rearing, respiratory and gastrointestinal diseases are major causes of illness and death. Although a number of antibiotics are available for use, few data on the relative frequency of treatment with these antibiotics in the veal industry are available (Sargeant and others 1994). Milk-based liquid starter diets fed to veal calves usually contain antibiotics for disease prophylaxis, until about 4 to 6 weeks of age when they are fed a milk-based liquid grower diet that does not contain an antibiotic (NRC 1999).

Minor species (sheep, goats, and bison). In the United States, minor species are defined by exclusion, as animals other than cattle, horses, swine, chickens, turkeys, dogs, and cats. In January 2005 the U.S. inventory of sheep and lamb totaled 6.14 million head (2.84 million slaughtered for food use), compared with cattle and calf inventory in July 2004 of 103.6 million (USDA 2005a, 2005b).

Six antibacterial drugs are approved for use in sheep, one of which–chlortetracycline–is approved for growth promotion and feed efficiency (NRC 1999). The focus of antibiotic treatment in sheep is the prevention and control of respiratory diseases, including shipping fever. Methods for administering drugs to sheep flocks include incorporation into feed or water, injection, and oral dosing. Treatment methods in goats are similar to those in sheep except that goats tolerate oral drenching less well, and in the United States it is common for goats to be treated as individuals rather than as herds. As ruminants, these species also receive protocols for the prevention and treatment of mastitis. Two antibiotics, neomycin and penicillin/streptomycin, are approved for use for enteritis and various infections, respectively; four drugs are approved for use for coccidiosis and parasites (NRC 1999).

Currently there are approximately 350000 head of bison in North America (NBA 2005); about 30000 head were slaughtered for food production in the United States in 2004 (USDA 2005a). Use of antibiotics in bison production is generally discouraged; and occurs only for treatment purposes. The Source Verification Program of the National Bison Association, which provides the standards for "certified buffalo products," prohibits administration of low doses of antibiotics over a long period of time. Medicated feeds are only permitted at "treatment" levels prescribed by a veterinarian.

While there are several other minor species used for food production, because their contribution to antimicrobial resistance is relatively small they are outside the scope of this report.

Food Animal Slaughter

Food animal slaughtering facilities in the United States apply carcass sanitization or physical or chemical decontamination treatments immediately before and after hide removal, at the end of the dressing process (before carcass chilling), and potentially after chilling; EU regulations, however, do not allow use of chemical decontamination agents in slaughter facilities (Koutsoumanis and others 2006; Sofos 2002b; Stopforth and Sofos 2006).

Some chemical agents may be incorporated into cleaning or washing solutions to reduce hide contamination. Solutions evaluated for this purpose include cetylpyridinium chloride, lactic acid, sodium hydroxide, ethanol, trisodium phosphate, acidified chlorine, and phosphoric acid. Chemical dehairing is a patented cattle hide decontamination process that involves use of a sodium sulfide solution followed by neutralization with hydrogen peroxide. This process is expected to minimize the importance of animal hides as sources of environmental and carcass contamination (Sofos and Smith 1998; Stopforth and Sofos 2006).

A number of interventions exist for sanitizing or decontaminating carcasses or fresh meat and poultry in the United States. These include water or steam (that is, hot water, pressurized steam, steam-vacuum) and chemical solutions, especially organic acids. These interventions significantly reduce bacterial populations, including those of enteric pathogens such as Escherichia coli O157:H7 and Salmonella. Such bacterial reductions allow the industry to meet regulatory (USDA/FSIS 1996) and contractual criteria. Spraying or rinsing of carcasses with an organic acid solution (for example, lactic and acetic acids) before evisceration and chilling reduces total bacterial populations and pathogen prevalence, and may also result in residual antimicrobial activity during product storage (Koutsoumanis and Sofos 2004a; Koutsoumanis and others 2004; Sofos and Smith 1998). Although this intervention reduces the prevalence and probably the concentration of E. coli O157:H7 on meat carcasses, concern has been raised that the treatment may select for, lead to adaptation of, or enhance the inherent tolerance of pathogen cells to acid (Samelis and Sofos 2003a, 2003b). In vitro studies have indicated the potential for sublethal organic acid rinsing treatments, which depend upon pH, acid type, and exposure duration, to cause acid stressing and selection of acid-resistant survivors in fresh meat decontamination runoff fluids. A potential concern is that any survivors may create niches in the plant environment for cross-contaminating subsequent batches of fresh meat (Samelis and others 2001a, 2002a, 2002b, 2003, 2004a, 2005b).

Additional chemical solutions for fresh meat and poultry decontamination include chlorine-based compounds and trisodium phosphate, which are used in the poultry industry, and acidified (usually with citric or lactic acid) sodium chlorite, hydrogen peroxide, ozonated water, activated lactoferrin, and peroxyacetic acid-based preparations. A variety of other tested chemical compounds such as polyphosphates, benzoates, propionates, sodium hydroxide, sodium metasilicate, and sodium bisulfate have shown various rates of success for decontaminating meat and poultry (Sofos 2002b; Stopforth and Sofos 2006).

Within a multiple hurdle approach to microbial control, fresh meat decontamination may involve the simultaneous sequential application of treatments that act synergistically or additively. Described by Leistner and Gould (2002) and Sofos and Smith (1998), a hurdle technology approach is the application to food of multiple physical, chemical, and biological antimicrobial factors at individually sublethal levels, rather than as a single hurdle at a higher, lethal level. When used in proper combinations, sublethal levels of antimicrobials are adequate for pathogen control, that is, microbial inactivation or growth inhibition. The multiple hurdles are designed to collectively lead to pathogen inactivation through metabolic exhaustion or growth inhibition for a certain period of time (Leistner and Gould 2002). For example, in fresh meat decontamination, the multiple hurdle approach may involve the simultaneous (for example, warm acid solutions) or the sequential (for example, hide cleaning, carcass steam vacuuming, pre-evisceration carcass washing, hot water, steam treatment, and organic acid rinsing treatments before carcass chilling, spray chilling of carcasses, and post-chilling-before-boning chemical treatments) application of treatments (Stopforth and Sofos 2006).

Effectiveness of hurdles may depend on the number and type of treatments, their intensity, and application sequence. For example, lactic acid rinsing of beef after hot water washing is more effective for microbial reduction and, especially, control of microbial growth during storage than before hot water washing (Koutsoumanis and Sofos 2004a; Koutsoumanis and others 2004; Koutsoumanis and others 2006; Sofos and Smith 1998). Synergism of an acid-heat-dehydration hurdle system was shown effective for inactivating E. coli O157:H7, Salmonella and Listeria monocytogenes inoculated pre- or post-drying on beef subsequently used to produce jerky, a North American dried meat snack (Calicioglu and others 2002a, 2002b, 2003a, 2003b, 2003c, 2003d; Yoon and others 2005). Selection of hurdles, their intensity, and sequence of application should aim at maximizing control without pathogen stress-adaptation or selection of resistant cells (Samelis and Sofos 2003a).

Aquaculture. Various types of aquaculture involving many different food-fish species are practiced worldwide. The extensive type of aquaculture practiced before 1980 has given way to more intensive pond, cage, net-pen, raceway (flow-through), and closed recirculating system culture. In the year 2000, salmon, tilapia and hybrid striped bass production in the United States reached 49 million, 20 million, and 10 million pounds, respectively (Carlberg and others 2000; Posadas 2003a, 2003b). Total production of channel catfish reached 630 million pounds and the rainbow trout industry produced approximately 46 million pounds of trout 12 inches or larger in 2003 (NWAC 2003; USDA 2004a).

Increase in demand and production capability has led to an increased concern about diseases, especially bacterial diseases. Antibiotics are only approved to treat disease as labeled and cannot be used in aquaculture prophylactically or for growth promotion. Antibiotics are incorporated into medicated feeds and are never added to the water to treat bacterial disease. Management and control of bacterial diseases are accomplished by administering medicated feeds or vaccines, and implementing improved husbandry practices. In the United States, only four antibiotics, Romet® (sulfadimethoxine/ormetoprim 5:1, Hoffman LaRoche, Nutley, N.J., U.S.A.), Terramycin® (oxytetracycline, Pfizer, Inc., U.S. Animal Health Operations, New York, N.Y., U.S.A.), sulfamerizine (no longer manufactured or available for aquaculture use), and Aquaflor® (florfenicol, Shering Plough Animal Health, Kenilworth, N.J., U.S.A.) are approved for use in aquaculture. Antimicrobials used in the U.S. aquaculture industry are regulated by the FDA.

In the United States, production of channel catfish, Ictalurus punctatus, is the largest and most economically important form of intensive aquaculture. A typical catfish farm contains brood-fish holding and spawning ponds, a hatcher, fingerling nursery, and grow out ponds. The ponds are earthen-bottomed and typically 10 to 20 acres in size. Catfish production has increased from 500 fish/acre in the industry's infancy to current levels of 10000 fish/acre. Of the losses caused by infectious disease in food-size channel catfish, approximately 60% are the result of single or mixed infections of Edwardsiella ictaluri, the causative agent of enteric septicemia of catfish (ESC), and Flavobacterium columnare, the causative agent of columnaris disease (Khoo 2001). Most catfish farmers are familiar with the clinical signs of the common bacterial diseases of catfish, and at the first sign of disease, a sample of sick fish is collected and shipped to the nearest aquatic diagnostic laboratory. Diagnostic laboratories typically culture the causative agents of disease and perform antibiotic susceptibility testing on bacterial pathogens.

Antibiotic use in catfish culture escalated in 1981 with the emergence of ESC, until approximately 1997, when management trends began to change. Sulfadimethoxine/ormetoprim 5:1 has traditionally been the most popular drug premix, because it is incorporated into a floating feed, but this situation may change with the approval in 2005 of florfenicol medicated feed. Oxytetracycline is only available in a sinking feed, which is less desirable because feeding activity is difficult to monitor (MacMillan 2003).

The practice of stocking and growing tilapia and hybrid striped bass at very high densities in closed recirculating aquaculture systems has led to the emergence of several bacterial pathogens, most notably Streptococcus iniae, as a limiting factor in production. Cumulative mortality rates in young fish can reach 75% in a matter of weeks although mortality is usually not as explosive as for other bacterial diseases of fish (Plumb 1999). Currently no antibiotics are approved by FDA for treating bacterial diseases in tilapia or hybrid striped bass.

The rainbow trout industry has greater maturity than many other forms of aquaculture and benefits from years of research on the diseases of salmonids and best management practices for those diseases. Many large trout producers have their own staff of fish pathologists who are responsible for maintaining the health of the fish stocks. The most prevalent bacterial diseases of rainbow trout are enteric redmouth disease (ERM) caused by Yersinia ruckeri, bacterial kidney disease caused by Renibacterium salmoninarum, furunculosis caused by Aeromonas salmonicida and coldwater disease caused by Flavobacterium psychrophilum. Asymptomatic carriers are common with ERM resulting in efficient disease spread. Once considered a major problem in the farm-raised trout industry, ERM is largely controlled today by good management practices and vaccination, although oxytetracycline medicated feeds have also been successfully used. Oxytetracycline medicated feeds have been used successfully at 50—75 mg/kg of fish/day for 10 days followed by a 21-day withdrawal period (Plumb 1999). ERM was one of the first fish diseases to be managed by vaccination. Current practices involve vaccination of 4 to 4.5 g fingerlings by immersion in a killed bacterin (suspension of killed or attenuated bacteria for use as a vaccine), which provides protection for 12 mo. The success of the ERM vaccine has resulted in greatly reduced mortality, reduced antibiotic usage, and reduced feed conversion rates in U.S. rainbow trout (Plumb 1999). Bacterial kidney disease remains difficult to manage, and is currently treated by chemoprophylaxis by injecting brood stock with 20 mg/kg erythromycin. Management of furunculosis involves the use of disease resistant strains of fish, destruction of infected fish, facility sanitation, and restrictions on use of eggs from infected broodstock. Sulfadimethoxine/ormetoprim medicated feeds have been used successfully at 50 mg/kg of fish/day for 5 d with a 42-d withdrawal period. Vaccines have not been as successful commercially because injectable vaccines are required to elicit adequate protection.

In salmonid mariculture (cultivation of marine organisms in their natural environment), vibriosis has been implicated as an important disease. Several species of vibrio bacteria, particularly Vibrio anguillarum, V. ordalii, and V. salmonicida, are responsible for the disease. Oxytetracycline has been used to treat vibriosis with variable results. Bivalent vaccines with antigenic components from V. anguillarum and V. ordalii are currently used with great success.

In the early 1990s, several mariculture ventures were established in brackish-water areas of south Louisiana where hybrid striped bass, Morone saxatilis x M. chrysops, and red drum, Sciaenops ocellatus, were cultured in cages, net-pens, and ponds. The emergence of Photobacterium damselae subsp. piscicida as an important marine bacterial pathogen of hybrid striped bass, led to the use of antibiotic medicated feeds in an attempt to control mortality (Hawke and others 2003). Oxytetracycline, sulfadimethoxine/ormetoprim 5:1, and amoxicillin at 50 mg/kg fish/day were used to treat outbreaks of P. damselae subsp. piscicida in red drum and in hybrid striped bass on mariculture farms. The antibiotics were used after filing for permission from the FDA but were unsatisfactory for several reasons–poor efficacy due to rapid onset of disease and anorexia of sick fish, recurrent infections following the use of antibiotics, and rapid development of antibiotic resistant strains of P. damselae subsp. piscicida due to acquisition of R-plasmids (Hawke and others 2003).

In many instances, medicated feeds have not proven to be efficacious in aquaculture for a variety of reasons. Individual fish infected with bacterial diseases tend to go off feed early in an epizootic and will not receive a therapeutic amount of the antibiotic. For antibiotic feeds to effectively control an outbreak of disease, the majority of fish in the population must be actively feeding for individuals to receive a therapeutic dose. For this reason, early diagnosis and initiation of therapy are paramount. Additionally, maintenance of good water quality and parasite control are important to keep feeding responses high.

Plant agriculture. The types of antimicrobials used in plant agriculture include antibiotics for control of certain bacterial diseases, and fungicides for control of fungi. Fungi and viruses are the most prevalent microorganisms causing diseases of plants; bacteria are relatively minor in importance, with some notable exceptions. Fruit trees account for most antibiotic use on plants in the United States (McManus 2000). In the United States, streptomycin and oxytetracycline have been used for more than 40 y as preventative treatments to control bacteria, primarily, affecting fruits and vegetables. Trees are generally sprayed during blossom time, when they are most susceptible to infection by Erwinia amylovora (causal agent of fire blight) and Pseudomonas syringae pathovar papulans (causal agent of apple blister spot). The edible fruit is not sprayed. Although streptomycin is registered by the EPA for use on 12 fruit, vegetable, and ornamental fruit crops, and oxytetracycline is registered for use on 4 fruit crops (Vidaver 2002), a limited number of fruit tree species–apple, pear, and peach–are treated in such a manner by antibiotics.

Most antimicrobials used in plant agriculture are fungicides. The top 12 economically severe fungal diseases are: cereal rusts, cereal smuts, ergot of rye and wheat, late blight of potato, brown spot of rice, southern corn leaf blight, powdery and downy mildews of grapes, downy mildew of tobacco, chestnut blight, Dutch elm disease, and pine stem rusts. Some of these diseases are worldwide and some are more restricted, due to host and climate (Agrios 2005).

Of the approximately 135 fungicides in 40 chemical classes (FRAC 2003), a large number are chemically classified as azoles. These popular fungicides are relatively cheap, have broad spectrum systemic activity for both preventative and curative effects, and are relatively stable (Hof 2001). The azoles are effective against mildews and rusts of grains, fruits, vegetables, and ornamentals; powdery mildew in cereals, berry fruits, vines and tomatoes; leaf spots and flower blights in flowers, shrubs and trees, and several other plant pathogenic fungi (Hof 2001). At present, there are no cross-over chemicals with those used in human medicine to treat serious systemic mycoses. However, although the formulations differ in their imidazole or triazole ring or in the side chain, in all cases the fungal target site (the enzyme lanosterol 14α-demethylase) is the same (Dismukes 2000). Fungicide resistance in plant pathogens may be of concern to those treating medical mycoses.

Residues of antibiotics and fungicides on fruits and vegetables are monitored by the Environmental Protection Agency (EPA); the residues have not been considered of concern with respect to antimicrobial resistance. Treated microorganisms, however, may be present on fruit and produce. Thus, antimicrobial resistance of plant pathogens and resistance of microbes in the treated environment raise questions about the potential for compromise in the use of these antimicrobials in human disease treatment.

Genes coding for antibiotic resistance have been used as markers in transgenic plant production, which is used to indirectly recover the desired trait(s), that is trait(s) not previously achievable through conventional plant breeding. Thus, a desired trait from an unrelated plant, animal, or microbial source may be added to a plant's replication machinery in single-cell technology, but the transformed cells may not be selectable directly when grown as tissue culture in vitro. After initial indirect selection, some markers can be eliminated as the plant is allowed to grow normally. These recombinant DNA derived plants have raised questions about the potential transfer of antibiotic resistance to animals or humans, although there has been no conclusive evidence of gene transfer from plant chromosomes to animals or humans. The risk of transfer of antibiotic resistance markers and the corresponding hazard was reviewed by Bennett and others (2004), and found to be "remote" and "slight." Nevertheless, under the impetus of the EU, genes expressing resistance to antibiotics used in medical or veterinary treatment as markers will be phased out between 2004 and 2008.

Food processing

Several different types of antimicrobial agents (Tables 3 and 4) are used in food manufacturing to either clean or sanitize to prevent cross-contamination in food processing facilities, or ensure food quality and safety. Food antimicrobials were traditionally used to prevent food spoilage, and only recently have been applied to control pathogen growth. Unlike the approval process for use of antibiotics in animals, which requires a risk assessment of resistance acquisition, the potential for the development of resistance to food antimicrobial agents is not considered during their approval for food use.

Table 3–Sanitizers commonly used in the food industry (Davidson and others 2005; McDonnell and Russell 1999)

Active ingredientEnvironmental surfacesFood contact surfacesFood tissuesRestroomHandcare
Alcoholsa + +
Oxidizing compoundsb + +
Hypochlorite + + +
Quaternary ammonium compounds + + ± +
Phenolics +
Acid anionics + +
Acidified sodium chlorite ± ± +
Chlorine dioxide + + +
Triclosan +
Para-chloro-meta-xylenol +
Chlorhexidine +

aIncludes ethyl alcohol (ethanol, alcohol), isopropyl alcohol (isopropanol, propan-2-ol), and n-propanol.
bIncludes hydrogen peroxide and peracetic acid.

Table 4–FDA-approved food antimicrobials (IFT 2002a)

Compound(s)Microbial targetPrimary food applicationsTitle 21 CFR designationa
Acetic acid, acetates, diacetates, dehydroacetic acid Yeasts, bacteria Baked goods, condiments, confections, dairy products, fats/oils, meats, sauces 184.1005, 182.6197, 184.1754, 184.1185, 184.1721, 172.130
Benzoic acid, benzoates Yeasts, molds Beverages, fruit products, margarine 184.1021, 184.1733
Dimethyl dicarbonate Yeasts Beverages 172.133
Lactic acid, lactates Bacteria Meats, fermented foods 184.1061, 184.1207, 184,1639, 184.1768
Lactoferrin Bacteria Meats b
Lysozyme Clostridium botulinum, other bacteria Cheese, casings for frankfurters, cooked meat and poultry products 184.1550c
Natamycin Molds Cheese 172.155
Nisin Clostridium botulinum, other bacteria Cheese, casings for frankfurters, cooked meat and poultry products 184.1538d
Nitrite, nitrate Clostridium botulinum Cured meats 172.160, 172.170, 172.175, 172.177
Parabens (alkyl esters (propyl, methyl, heptyl) of p-hydroxybenzoic acid) Yeasts, molds, Gram-positive bacteria Beverages, baked goods, syrups, dry sausage 184.1490, 184.1670, 172.145
Propionic acid, propionates Molds Bakery products, dairy products 184.1081, 184.1221, 184.1784
Sorbic acid, sorbates Yeasts, molds, bacteria Most foods, beverages, wines 182.3089, 182.3225, 182.3640, 182.3795
Sulfites Yeasts, molds Fruits, fruit products, potato products, wines Various

aFood and Drug Administration designations in Title 21 of the Code of Federal Regulations. Food antimicrobials approved by the U.S. Department of Agriculture's Food Safety and Inspection Service for use in meat products are listed in sections 424.21 and 424.22 of Title 9 of the CFR.
bFDA/CFSAN GRAS notice 000067, Oct. 2001.
cFDA/CFSAN GRAS notice 000064, Apr. 2001.
dFDA/CFSAN GRAS notice 000065, Apr. 2001.

Cleaning and sanitation. Equipment surfaces and the surrounding environment inevitably become soiled and require cleaning during food processing. In addition to detergents and soaps, antibacterial agents (biocides) are used as sanitizers, disinfectants, and handcare products throughout the food system. These substances are used to reduce the level of microorganisms on food contact surfaces, in food formulations, on ready-to-eat (RTE) food product surfaces, environmental surfaces, food tissue surfaces, and human skin. Formulations for these uses contain one or more antibacterial agents, commonly referred to as active ingredients, as well as other components including surfactants, pH buffering agents, and water conditioning agents. The active ingredients of sanitizers and various common uses in the food industry are shown in Table 3. Overviews of the cellular targets and inactivation mechanisms of biocides are provided in Figure 3 and 4.


Microbial inactivation and resistance to biocides. Reprinted with permission from the American Society for Microbiology (ASM News, January 2002, p 20–24).
Figure 3–Microbial inactivation and resistance to biocides. Reprinted with permission from the American Society for Microbiology (ASM News, January 2002, p 20—24).
Mechanisms of inactivation by biocides. Thiol and amino groups in all microorganisms are susceptible to the appropriate agents shown. In vegetative bacteria and fungi, ribosomes and DNA are susceptible to hydrogen peroxide and iodine and to acridine dyes, respectively. Reprinted with permission from the American Society for Microbiology (ASM News, January 2002, p 20–24).
Figure 4–Mechanisms of inactivation by biocides. Thiol and amino groups in all microorganisms are susceptible to the appropriate agents shown. In vegetative bacteria and fungi, ribosomes and DNA are susceptible to hydrogen peroxide and iodine and to acridine dyes, respectively. Reprinted with permission from the American Society for Microbiology (ASM News, January 2002, p 20—24).

Detergents may be classified into inorganic alkali (sodium hydroxide and sodium carbonate, for example), inorganic and organic acids (phosphoric and citric acids, for example), surface-active agents (for example, synthetic detergents–either anionic, cationic, non-ionic, or amphoteric [capable of reacting chemically as either an acid or base]), and sequestering agents (polyphosphates, ethylenediamine tetra acetic acid [EDTA, for example]). Modern detergents contain a mixture of different chemicals, each contributing to the desired properties of the formulation. Sanitizers used in the food industry can be classified into chlorine releasing compounds, QACs, iodophors10 and amphoteric compounds.

Quality and safety. Sanitization or decontamination treatments, similar to those applied on raw beef, may also be used for fresh produce (Beuchat and Ryu 1997; Taormina and others 1999). Combinations of thermal (hot water or steam) and chemical interventions (organic acid solutions) in the form of sprays or rinses are used successfully as sanitizing or decontaminating treatments on fresh produce to reduce overall microbial contamination and prevalence of pathogenic bacteria (Sofos 2002b; Sofos and Smith 1998; Stopforth and Sofos 2006).

Processing and preservation technologies involving manipulation of physical, chemical, and biological factors are used, often in combination, by food processors. The objective of their use is to ensure the stability and safety of foods by inactivating or inhibiting growth of spoilage and pathogenic microorganisms. For example, various combinations and sequences of sublethal hurdles in RTE meat and poultry products may also be applied to control post-lethality processing contamination with L. monocytogenes during product storage (see side bar), as required by new U.S. Dept. of Agriculture Food Safety and Inspection Service regulations (USDA/FSIS 2003).

Chemical preservatives and treatments. While some chemical food preservatives, such as common table salt, nitrites, and sulfites, have been in use for hundreds of years, most others have been extensively applied only in recent decades. Food preservatives used to prevent food deterioration caused by microbial growth are termed "food antimicrobial agents." The historical function of food antimicrobial agents is inhibition of spoilage microorganisms and extension of shelf life. The use of food antimicrobial agents to control pathogens is more recent and is increasing (Davidson and Zivanovic 2003). Food antimicrobial agents are generally not used alone to control foodborne pathogens, but are included as components of the multiple hurdle approach to microbial control. Exposure of E. coli O157:H7, Salmonella, or L. monocytogenes—inoculated apple slices or other produce to ascorbic and citric acid solutions, for example, enhanced destruction of the pathogens during subsequent drying (Burnham and others 2001; Derrickson-Tharrington and others 2005; DiPersio and others 2003, 2004; Yoon and others 2004). Other common applications of food antimicrobials include use of sodium nitrite to inhibit Clostridium botulinum in cured meats if product temperature abuse occurs, organic acid solutions as spray sanitizers to control pathogens on beef carcasses, nisin and lysozyme to control C. botulinum in pasteurized process cheese, and lactate and diacetate for L. monocytogenes control in processed RTE meat and poultry products (USDA/FSIS 2000).

Naturally occurring antimicrobials. Food antimicrobial agents may be classified as traditional or naturally occurring (Davidson 2001). Traditional food antimicrobial agents, listed in Table 4, undergo review and approval for food use by many international regulatory agencies. Naturally occurring antimicrobials, however, which include compounds from microbial, plant, and animal sources (Table 5) are limited in approval and applications (Sofos and others 1998). Nisin, natamycin, lactoferrin, and lysozyme are among the few naturally occurring substances that are approved by regulatory agencies in some countries for direct application to foods.

Table 5–Naturally occurring food-related antimicrobials and sources

Avidin Eggs Binds vitamin biotin
Chitosan Shellfish, mushrooms, fungi Aminoglycoside; interaction with cell wall polysaccharides or cytoplasmic membrane resulting in altered permeability or transport
Conalbumin (ovotransferrin) Eggs Iron chelation
Lactoferrin Milk Iron chelation; alteration of membrane permeability; prevention of binding
Lactoperoxidase Milk With H2O2 and hypothiocyanate forms inhibitor
Lysozyme Eggs, milk, biological secretions Catalyzes hydrolysis of 1,4-glycosidic linkages of peptidoglycan of bacterial cell walls
Caffeine, theophylline, theobromine Coffee, cocoa, tea Variable activity
Flavonoids (chalcones, flavones, flavonols, flavanones, anthocyanins, isoflavonoids) Plants Variable activity
Humulon(e)/lupulon(e) Hops Some activity against Gram-positive bacteria and fungi
Isothiocyanates Brassicaceae (Cruciferae)– mustard family Allyl isothiocyanate, horseradish extract; activity may be due to enzyme inhibition
Phenolic/hydroxycinnamic acids Plants Caffeic, p-coumaric, ferulic, chlorogenic, protocatechuic, vanillic, gallic
Oleuropein Olives Phenolic glycoside; cytoplasmic membrane disruption
Tannins Plants Hydrolyzable, condensed (proanthocyanidins)
Terpenes/terpenoids Spices Eugenol, thymol, carvacrol, cinnamic aldehyde, vanillin, pinene, camphor, citral, borneol, thujone, menthol; interaction with the cell membrane
Thiosulfinates Allium (onions, garlic) Inhibition of sulfhydryl containing enzymes
Bacteriocins Lactic acid bacteria Lactococcus, Pediococcus, Lactobacillus, Leuconostoc, Carnobacterium and others; bind to and form pores in cytoplasmic membrane
Natamycin Streptomyces natalensis Macrolide antifungal antibiotic; complexes with sterols in fungal cell membranes, disrupts cell membrane

Food biopreservation uses natural or controlled microflora and/or their antibacterial metabolic end products to interfere with undesirable microorganisms. Lactic acid bacteria (LAB), for example, occur either in the initial natural microflora of fermented or other foods or are added as starter cultures, where their growth dominates over that of other microbes during fermentation or retail case display and home refrigeration (vacuum-packaged meat, for example). Growth of LAB interferes with spoilage and pathogenic bacteria through nutrient and oxygen depletion, and production of inhibitory metabolic substances such as lactic and acetic acid, acetoin, diacetyl, hydrogen peroxide, reuterin, and bacteriocins (Koutsoumanis and Sofos 2004a; Koutsoumanis and others 2006).

Controlling L. monocytogenes in Ready-to-Eat (RTE) Foods

Recalls of RTE meat and poultry products and foodborne illness outbreaks involving fatalities attributed to L. monocytogenes led to the establishment of a new regulation for controlling the pathogen in meat and poultry products that may become contaminated after processing, during slicing and packaging, and in which their growth may be supported during product distribution and storage (USDA/FSIS 2003). According to the regulation, manufacturers of sensitive RTE meat and poultry products should select one of three alternative approaches for preventing contamination and inactivating or controlling the pathogen's growth during storage. In addition to physical processes (for example, heat, high hydrostatic pressure), the alternatives may be based on chemical compounds applied as antimicrobial agents or sanitizers. Substances such as potassium or sodium lactate, sodium acetate, sodium or potassium diacetate, nisin, acetic acid, lactic acid, sodium or potassium benzoate or sorbate, acidic calcium sulfate, and buffered citrate applied as formulation ingredients or postprocessing solutions are effective against the pathogen in such RTE meat and poultry products. The most common approach for controlling L. monocytogenes in RTE meat and poultry products combines sodium or potassium lactate with sodium diacetate in the product formulation (Tompkin 2002). Alternative antimicrobial approaches may be based on combinations of physical and chemical antimicrobial hurdles applied as formulation ingredients during processing, or as postlethality treatments, including spraying or dipping solutions during packaging (Barmpalia and others 2004, 2005; Bedie and others 2001; Geornaras and others 2005; Samelis and others 2001c, 2002c, 2005a).

Home products

Antimicrobials are increasingly more commonplace in consumer products for home use. Levy (2001) reported that more than 700 antibacterial-containing products (for example, cleansers, soaps, toothbrushes, dishwashing detergents, hand lotions, plastic food storage containers, and bedding and bedding linens) were being marketed for the home. Other uses include food contact surfaces (cutting boards, for example), environmental surfaces, personal hygiene products, and food tissue antimicrobial sprays. Triclosan (TCS; 2,4,4'-trichloro-2'-hydroxydiphenylether), for example, has been used in skin-care products (soaps, for example) for some 30 y, and has also been used in handwashes and dental hygiene products (Russell 2004). Triclosan and parachlosulfadimethoxine/ormetoprimaxylenol (PCMX) are the most common antimicrobials used in antimicrobial hand soaps. Triclosan has also recently been incorporated into plastics such as cutting boards and knife handles, which are used in both institutional and industrial settings (Bhargava and Leonard 1996). This broad-scale use has prompted widespread concerns over the development of resistant organisms.

Human medicine

Antibiotics are used in humans in community and hospital settings primarily to treat disease, but are also used to prevent infection. The activity, action, and resistance of antiseptics and disinfectants used in hospitals and other health care settings for a variety of topical and hard-surface applications were reviewed by McDonnell and Russell (1999).

Quantitative Usage Data

For the reasons addressed below, it is difficult to estimate how antibiotic usage is distributed among human, veterinary, and plant applications, and the exact amount of antibiotics introduced annually into the environment. Although the exact portion of antibiotics used in production agriculture is unknown, it is certainly significant, and likely comparable to the amount used in human medicine.


An Institute of Medicine (IOM) Committee on Human Health Risk Assessment of Using Subtherapeutic Antibiotics in Animal Feeds attempted to quantify the use of antibacterial agents in livestock and poultry feeds (IOM 1989). Using International Trade Commission (ITC) data from 1950 onward, the IOM committee estimated that in 1985 total production of antimicrobials was 31.9 million pounds. The committee noted that the reliability of the production data used in the analysis was unknown. It was estimated that 16.1 million pounds were used for disease prevention and growth promotion in animals and that 2.3 million pounds were used for disease treatment. Comparable data have not been available from the ITC since 1986, preventing updates of this estimate. Although these often-cited figures are no longer current, they provide a benchmark and demonstrate one method for quantifying antimicrobial usage.

The Union of Concerned Scientists (UCS) derived antimicrobial use estimates for cattle, swine, and poultry on the basis of drug label indications, estimates of herd size, and extent, intensity, and duration of use in each commodity or sector (Mellon and others 2001). The UCS integrated data from the USDA, National Research Council, National Animal Health Monitoring System (NAHMS), and IOM to estimate that 24.6 million pounds of antimicrobials were used for nontherapeutic uses (defined by UCS to include uses for prevention and control of disease as well as for growth promotion) in cattle, swine, and poultry in 1999. Criticisms of the UCS method of assessment included their assumptions, which were: (1) uniformity of production conditions; (2) lack of variation in use practices across producers due to product cost or personal preference; and (3) constant herd/flock size from 1984 to the late 1990s (Jones and Ricke 2003).

More recent estimates of antimicrobial usage are available from the Animal Health Institute (AHI), which estimates antibiotic use (including ionophores and arsenicals) in farm and companion animals from data comprising responses to surveys of AHI member animal health companies. The surveys ask for the total quantity of active ingredients manufactured and sold in a calendar year by drug class, and the estimated percentage sold for the purpose of therapeutic and health maintenance (as measured by improved growth rates or more efficient feed use). AHI estimates show a general downward trend in total antibiotic use between 1999 and 2004. Production decreased from 24.9 million pounds in 1999 (of which 88.3% was for therapeutic use), to 22 million pounds in 2002 (of which 91% was for therapeutic use), and to 21.7 million pounds in 2004 (of which 95% was for therapeutic use) (AHI 2002, 2004, 2005). The AHI estimates do not include all quantities of generic antibiotics because many manufacturers of generic drugs are not AHI members. Since the majority of antimicrobials used for growth promotion are approved for other indications as well, it is difficult to determine how they were categorized by the survey respondents.

Although the AHI and UCS estimates for total use appear similar, the AHI production estimates include total animal use for all species and indications. The UCS estimates included solely nontherapeutic use in only the three major food animal species―beef cattle, swine, and broiler chickens. Thus, the UCS estimate for nontherapeutic antimicrobial use in a limited number of species is roughly 10 times the AHI estimate for all species. As noted, the UCS categorization of drugs having multiple approved uses is unclear, further complicating the interpretation of the figures. Additional points relevant to the AHI and UCS estimates are: (1) UCS used the term nontherapeutic to include prophylactic and growth promotion uses in only food animals, while AHI included growth promotion and therapeutic uses among all animals, including companion animals; (2) UCS estimated the percentage of a given food animal population that was medicated and multiplied by the product's label dosage; however, some approved products were never marketed, and others are used at varying dosage rates; (3) AHI used data provided by companies on their marketed products; other than an estimate of antimicrobials used for growth promotion, no attempt was made to further characterize usage per animal species nor to factor in the dose or duration of use; AHI did not include generic usage data; UCS may have; (4) AHI combined products into groups of antibiotics to comply with anti-trust regulations of trade associations. Therefore, despite the AHI and UCS estimates, reliable data on the amount of antibiotics used are not available, which makes assessment of effects and management difficult.

In 1999, the Alliance for the Prudent Use of Antibiotics (APUA) initiated the multidisciplinary Facts about Antibiotics in Animals and the Impact on Resistance (FAAIR) Project, which identified the critical gap in surveillance data on antimicrobial use in animals and recommended that such data be made available to improve risk assessment and better inform policy decisions on antimicrobial use in animals (FAAIR 2002). Although the World Organization for Animal Health (OIE) has proposed guidelines for the collection of quantitative antibiotic usage data, a standard method for assessing use has yet to be applied (OIE 2004). Following up on FAAIR, APUA established the Advisory Committee on Animal Antimicrobial Use Data Collection in the United States to determine the most effective means for gathering data on antimicrobial use in food animals. Comprised of varied stakeholders, from academia, government, the food animal production sector, the animal health industry, human health industry, public interest organizations, research community, and veterinarians, the committee identified methodological options for data collection. Four major categories of antimicrobial use data were identified based on the source of information and its proximity to actual use–end-user data, prescription data, manufacturing data, and distribution data.

The Advisory Committee concluded that the ideal animal antimicrobial use data collection strategy would likely combine two or more of the methods identified by the committee. Because consensus could not be reached on the ideal combination of data methods, experts comprising the committee individually rated six methodological options. The methodological options are: (1) all practices/producers record all prescriptions/use indefinitely, (2) sentinel practices/farms track use electronically, (3) selected practices/producers record all prescriptions/uses for a defined period of time, (4) periodically survey a cross-section of veterinarians/producers, (5) solicit production and sales information from manufacturers, and (6) publicly disclose production information obtained by FDA from manufacturers (DeVincent and Viola 2006).


A survey conducted by the National Aquaculture Association to estimate the quantity of drugs used in the U.S. aquaculture industry indicated that only 22680 to 31750 kg of active antibiotic ingredients are sold per year (MacMillan and others 2003). Because of the small size of the U.S. aquaculture industry, and the fact that there is only one manufacturer of sulfadimethoxine/ormetoprim and one manufacturer of oxytetracycline, it is possible to accurately estimate the amount used on farms. From January 2001 to February 2003, 36126 kg of sulfadimethoxine/ormetoprim 5:1 and 22334 kg of oxytetracycline were sold for incorporation into medicated feeds for the aquaculture industry. Some minor use occurs when medicated feeds are purchased in Canada for use in U.S. salmon farms.


Fungicides are used more extensively on fruits than vegetables, with 99% of tart cherry acreage, 96% of table grape acreage, and 94% of land used for raspberry production receiving fungicidal treatment. Among vegetables, bulb onion, strawberry, and tomato led in fungicide applications, on a percent treated basis, with 87%, 86%, and 86% of acres treated, respectively (USDA 2004b). Fungicides are used much more extensively than antibiotics, with about 24000 metric tons (26000 tons) used in the United States per year.

The total amount of antibiotics used in plant agriculture has stayed fairly constant over the last decade (McManus and others 2002). In 2003, 7500 kg (16500 lb) of streptomycin were applied to about 15% of the apple and 32% of the pear acreage. Oxytetracycline use has increased from 7270 to 12270 kg (16000 to 27000 lb) between 1997 and 2003 (USDA 2004b), probably due to widespread streptomycin resistance of the target pathogens, especially on the East and West Coasts of the United States. The prevalence of imported produce necessitates an understanding of practices in the rest of the world, which in many cases are not known or reported.


Although estimates have been attempted, the quantity of human usage of antibiotics in the United States is unknown. Comprehensive estimates of total human use per annum in the United States have been reported by the AHI and UCS through their respective efforts to quantify antibiotic and antimicrobial use in food animals. AHI reported in 2000 that 32.2 million pounds of antibiotics are used annually in human medicine (AHI 2000). AHI obtained this figure indirectly by subtracting its estimate for total animal use (17.8 million pounds) from the 1989 IOM estimate of 50 million pounds (extrapolated from trends in the 1970s and 1980s to the 1990s) of use in both animals and humans (IOM 1989). The UCS estimate for human use (for inpatient and outpatient disease treatment and as topical creams, soaps, and disinfectants) was 4.5 million pounds. UCS estimates were based upon data compiled by the CDC National Center for Health Statistics (NCHS) survey of outpatient prescriptions and use, expert consultation, and a national market survey of inpatient hospital use (Mellon and others 2001).

Opportunities to acquire data on human use are greater than for animal use. In the United States, data are collected through several surveys conducted by the CDC's NCHS and the National Nosocomial Infections Surveillance (NNIS) System, comprising a collection of nosocomial (originating or taking place in a hospital) infection surveillance data from more than 300 hospitals. For the purpose of analysis, grams of antibiotics used are converted into the number of "defined daily dose(s)" (DDD) used each month in each hospital area. As defined by the World Health Organization (WHO), a DDD is an average daily dose in grams of a specific drug administered to an average adult patient (Ronning 1999). CDC also supports the collection of antibiotic use data through the Medication-Associated Adverse Event Module of the National Healthcare Safety Network (NHSN).

Private corporations are also sources of information. Under a 5-y contract established with the FDA in 2001, IMS Health (Fairfield, Conn., U.S.A.), an international corporation serving the pharmaceutical and healthcare markets with data sources from more than 29000 suppliers, has been providing market research information on drug use and the impact of pharmaceutical products on patient outcomes. The specificity and public availability of these data, however, are not yet known (IMS 2001).

Through the use of DDDs, it has been recently determined that antibiotic prescription rates within Europe vary markedly (Molstad and others 2002). In 2000, France and Germany consumed higher numbers of DDDs per capita, while the Netherlands and Denmark consumed fewer DDDs (Patrick and others 2004). In contrast to the United States, several countries, such as Denmark and Spain, have databases containing information on all antibiotics prescribed for all patients (Patrick and others 2004).

More recently, increasing attention has been given to the types of antibiotics being prescribed (Huang and Stafford 2002; Linder and Stafford 2001; Piccirillo and others 2001). Unlike the situation with animal usage, federal survey-based systems track human prescriptions and may serve as data sources for estimating use in human medicine. Antibiotic sales data are available from manufacturers, but there are limitations–sales data are not synonymous with actual consumption data, methodology is proprietary, production data are lacking.

During 2002 and 2003, penicillins were the most prescribed class of antibiotics in hospital outpatient and physician office visits in the United States (HHS/CDC/NCHS 2005). The number of antibiotic prescriptions in adults and children in U.S. ambulatory care settings declined from 151 million to 126 million between 1992 and 2000 (McCaig and others 2003). Also documented during this time period was evidence of increasing outpatient use of amoxicillin and cephalosporins (Steinman and others 2003). The 34% decrease in the rate of prescriptions written for children during physician office visits, and lack of increase for adults during a 20-y span may suggest that the efforts of the CDC, medical associations, and other stakeholder groups may be having a beneficial effect on prudent antibiotic use and overall prescription writing.

Factors contributing to the overuse of antibiotics in humans include real or perceived pressure from adult patients and parents of child patients to prescribe antibiotics, inadequate identification of label indications for some drugs, lack of awareness of prescription guidelines, the move toward managed healthcare, and inadequate time for physicians to explain to patients that antibiotics are often unnecessary (Hutchinson and Foley 1999; Okeke and others 1999; WHO 2002). A Congressional Research Service report noted that 96% of pediatricians surveyed reported that parents of children in office visits specifically requested an antibiotic prescription, and 33% prescribed an antibiotic without a clinical basis simply to appease the parent (Vogt and Jackson 2001). Hamm and others (1996) stated that parents and patients perceive that "they haven't gotten their money's worth" in appointments with primary care physicians that do not result in a prescription being written. Additionally, Avorn and Solomon (2000) pointed out that the number of patients seen per hour by physicians is increasing due to increasing administrative demands and that writing a prescription can serve as a termination strategy for an office visit.

Mechanisms for Emergence and Dissemination of Antimicrobial Resistance


As pointed out by Courvalin (2005), resistance to antimicrobial drugs is an unavoidable aspect of the general evolution of bacteria that occurs by chance. Mechanisms for emergence of bacterial resistance are quite diverse as are the modes of action of antimicrobials, which may include inhibition of various steps of DNA replication, transcription, and translation, or action at the level of the cell wall or cell membrane.

Microbial strategies for resisting the effects of antibiotics include impaired uptake, modification or overproduction of the target sites of antimicrobials, bypass of sensitive steps, absence of enzymes or metabolic pathways, and efflux of the antimicrobial drug (Russell and others 1997). Further, bacteria can resist the effects of antimicrobials by enzymatically degrading the drug before it reaches its target site, altering the protein(s) within the bacterium that serve as receptors for the antimicrobials, and changing their membrane permeability to the antibiotics (Cloete 2003; Dever and Dermody 1991).

Efflux pumps. Called "multidrug efflux pumps," these systems for transporting substances out of cells often provide resistance to a variety of structurally different antimicrobials, including antibiotics, dyes, and surfactants. Along with impaired uptake, efflux pumps are a main strategy that bacteria use to deal with the stress of sanitizer exposure (Russell and others 1997). Gram-positive and Gram-negative bacteria use the same efflux system for ethidium bromide and QACs. Tetracycline resistance in E. coli is at least partially due to an energy-dependent efflux mechanism (McMurry 1980), and a similar mechanism has been implicated in E. coli fluoroquinolone resistance (Cohen and others 1988, 1989; Hooper and others 1989). In addition, the genes for multiple antibiotic resistance in Pseudomonas aeruginosa may be on an efflux operon11 (Poole and others 1993). Efflux mechanisms, however, do not pertain to bacteriocins, which do not accumulate intracellularly.

Acid tolerance can be viewed in terms of efflux ability. The mechanism by which organic acids inhibit microorganisms involves passage of the undissociated form of the acid across the cell membrane lipid bilayer. Once inside the cell, the acid dissociates because the cell interior has a higher pH than the exterior. Protons generated from intracellular dissociation of the organic acid acidify the cytoplasm and must be extruded to the exterior. Yeasts develop resistance to sorbic and other organic acids via several mechanisms. They use the enzyme H+-ATPase along with ATP (adenosine triphosphate) energy to remove excess protons from the cell. Inhibition and/or inactivation of the yeasts may be due to eventual loss of cellular energy or inactivation of critical cellular functions due to low intracellular pH. To prevent energy depletion, a membrane protein may be induced for decreasing ATPase activity and thus conserve energy (Brul and Coote 1999). Exposure of Saccharomyces cerevisiae to sorbic acid induces a multi-drug resistance pump (membrane protein ATP-binding cassette transporter Pdr12 [Holyoak and others 1999; Piper and others 1998]), which confers resistance by mediating energy-dependent anion extrusion (Piper and others 1998). To circumvent the problem of extruded anions and protons reentering the cell upon recombining in the extracellular medium, adapted yeasts apparently reduce diffusion of the weak acids, most likely by altering cell membrane structures to reduce passage of the acids into the cell (Brul and Coote 1999). Similar mechanisms likely also exist for bacteria capable of developing resistance to sorbic or other organic acids.

Enzymatic degradation. A common phenomenon, enzymatic degradation, is the primary mechanism of resistance to β-lactam antibiotics via the hydrolysis of the β-lactam ring (Bush and Sykes 1984) and the resistance mechanism for chloramphenicol and aminoglycosides. Resistance to chloramphenicol, a broad spectrum antimicrobial, occurs through acetylation catalyzed by chloramphenicol acetyltransferase; other modes of resistance are also possible, however (Dever and Dermody 1991; Kucers and Bennett 1987). Methylases, acetyltransferases, nucleotidyltransferases, and phosphotransferaces are used against aminoglycosides (Davies 1994; Shaw and others 1993). Enzymic degradation of food antimicrobial agents can be specialized or general, but would be different from the enzymes that inactivate antibiotics. For example, some bacteria metabolize citric acid, rendering it ineffective against them. In contrast, many proteases inactivate bacteriocins in a nonspecific fashion. A nisin dehydroreductase conveys resistance by inactivating a nisin dehydro residue (Jarvis and Farr 1971).

Alteration of receptors. Alteration of specific receptor sites prevents proper target recognition. Resistance to nalidixic acid is most often due to mutations in gyrA and gyrB, the genes encoding the target proteins of the antibiotic. Resistance to ciprofloxacin is also associated with mutations in gyrA and gyrB (Heddle and Maxwell 2002; Hooper 1995; Tankovic and others 1996).

Membrane permeability change. The most common form of intrinsic resistance to antibiotics is due to membrane structure and composition, which can naturally act as a permeability barrier or undergo change through acquired resistance mechanisms, as in the case of Gram-negative bacteria. E. coli resistance to β-lactam antibiotics, for example, occurs upon replacement of the outer membrane OmpF porin by the narrower OmpC porin (Nikaido and others 1983) and in Staphylococcus epidermidis glycopeptide resistance may occur through over production of glycopeptide binding sites within the cell wall peptidoglycan (Sanyal and Greenwood 1993). Resistance to nisin can result from spontaneous genetic mutation (designated Nism) involving bacteriocin adsorption or membrane insertion, presumably causing loss of cell membrane fluidity and hindering nisin insertion (Nism cell membranes are more solid than those of the wild-type strain).

Membrane fluidity can play an important role in resistance of L. monocytogenes to antimicrobials (Juneja and Davidson 1993). L. monocytogenes cells grown in the presence of C14:0 or C18:0 fatty acids have higher phase transition (Tc) and increased resistance to four common antimicrobials than cells grown in the presence of C18:1, which have lower Tc and are more sensitive. It is assumed that the higher phase transition temperature of the membrane fatty acids prevents effective penetration of the pore-forming bacteriocin. Nisin-resistant C. botulinum also have altered membrane fatty acid composition that would increase their membrane rigidity (Mazzotta and Montville 1999).

Stress-adaptation, co-selection, cross-resistance, and cross-protection

Mechanisms exist whereby microorganisms that are resistant to one antimicrobial may become resistant to others (Yousef and Juneja 2003). Exposure to subinhibitory concentrations of an antimicrobial, for example, may activate intrinsic resistance mechanisms, thereby decreasing susceptibility of the microbe to the inducing agent and in tandem decreasing susceptibility to other, unrelated antimicrobials. In other instances, resistance to several antimicrobials having unrelated targets or modes of action may result from co-selection, which involves sequential linking of separate genes conferring resistance to different antibiotics, often on plasmids or integrons,12 and transferred together. Cross-resistance is the occurrence of resistance to antimicrobials because they have the same molecular targets. Cross-protection occurs when adaptation to one stress is associated with increased resistance to another, unrelated stress. Correlations among these mechanisms are seen in some cases, but the root causes of the dissemination of the resistance remain unknown.

Strains of E. coli resistant to thymol and eugenol (essential oils found in thyme and cloves, respectively) were found to be more resistant to chloramphenicol (Walsh and others 2003). Because stable resistance to the essential oil components was not readily detected, the authors denoted the increased resistance as "tolerance" (Walsh and others 2003). In contrast, methicillin-resistant Staphylococcus aureus, however, were found to be as sensitive to oregano essential oil and its components, carvacrol and eugenol, as methicillin-sensitive strains (Nostro and others 2004). Resistance to carvacrol, however, which is associated with changes in the cellular membrane, apparently does not confer resistance to other membrane-active compounds. Bacillus cereus adapted to carvacrol were demonstrated to be more sensitive to subsequent nisin exposure than nonadapted cells (Pol and others 2001).

Bacteria are able to produce stress response proteins when subjected to subinhibitory levels of stress (Yousef and Juneja 2003). A variety of situations can induce transcription and translation of stress response proteins, which convey increased resistance to a multitude of stressors. For example, exposure of E. faecalis to subinhibitory levels of sodium chloride, sodium dodecyl sulfate, and bile salts conferred a protective effect against heat compared to nontreated cells (Flahaut and others 1997). Heat shock proteins (HSP) comprise one of the most well-studied classes of stress response proteins, although the HSP levels do not correlate with the extent or persistence of protection (Jorgensen and others 1996; Mackey and Derrick 1990). HSP are typically regulated by sigma factors such as RpoS or RpoH, which are subunits of RNA polymerase.

Salmonella enterica serovar Enteritidis and L. monocytogenes first exposed to alkali are more resistant to heat treatment than those not pre-exposed (Humphrey and others 1991; Taormina and Beuchat 2001). Studies with Salmonella Enteritidis showed that treatment with low levels of alkali (pH 10.0 sodium hydroxide or trisodium phosphate) resulted in a decrease in protein expression of 15% and 22%, respectively (Sampathkumar and others 2004). Some outer membrane proteins, identified as protein chaperones and housekeeping proteins involved in biosynthesis, were up-regulated. Similarly, when E. coli K-12 was shifted from pH 7 to 8.8, known HSPs were induced (Taglicht and others 1987).

Hong and others (2002) found that Streptomyces coelicolor, containing a plasmid encoding a signal transduction system including the sigma factor E (ΦE), demonstrated lysozyme-induced resistance to kanamycin (100g/mL). L. monocytogenes has been shown to contain a similar signal transduction system (CesRK) that is activated upon introduction of lysozyme to the cells and results in antibiotic resistance.

An example of an intrinsic resistance system is the multiple antimicrobial resistance (mar) operon, a global regulator that controls intrinsic resistance to unrelated antibiotics and other cytotoxic substances (Alekshun and Levy 1999). Golding and Matthews (2004) demonstrated decreased susceptibility of E. coli O157:H7 to multiple antimicrobials, putatively linked to a mutation in the mar operon, following exposure to chloramphenicol. Potenski and others (2003) found that upon treating Salmonella Enteritidis cells with sublethal levels of chlorine, sodium nitrite, sodium benzoate, or acetic acid, the cells exhibited resistance to tetracycline, chloramphenicol, nalidixic acid, and ciprofloxacin, thus determining that a mar operon was responsible for the resistance responses.

Antimicrobial resistant phenotypes of E. coli O157:H7 may also be related to acquisition of class 1 integrons (Zhao and others 2001a), which is significant because the integrons may contain several antimicrobial gene cassettes and, therefore, co-select for resistance to other antimicrobials.

Genes encoding for multidrug efflux systems in S. aureus have been located on plasmids (generally 18 to 57 kb in size) also containing genes for resistance against penicillin, gentamicin, trimethoprim and kanamycin (Lyon and others 1984). The qacA and qacB genes have been found on plasmids that also confer resistance to various antibiotics, including penicillin (Lyon and Skurray 1987). Twenty-four QAC-resistant Staphylococcus isolates were analyzed for resistance to selected antibiotics and dyes (Heir and others 1999). Five of the seven strains with the QAC resistance genes qacA/qacB had high-level resistance to penicillin G and ampicillin. One isolate containing the smr gene showed resistance to ampicillin, penicillin G, tetracycline, erythromycin, and trimethoprim, but not to chloramphenicol, gentamicin, norfloxacin, kanamycin, or vancomycin. It was suggested that the antibiotic resistance in this strain was due to resistance markers on the chromosome or other plasmids harbored by the strain. All other sanitizer-resistant isolates were generally susceptible to antibiotics.

Several studies have found a lack of cross-resistance between agents, even when mechanisms appear similar. For example, when acquired resistance mechanisms for biocides, which can closely resemble those for antibiotics, were studied by Aase and others (2000), no connection was found between QAC resistance and antibiotic resistance in L. monocytogenes. They evaluated 200 L. monocytogenes isolates from various food, human, and environmental sources from Norway and Europe and found that 10% were resistant to benzalkonium chloride (BC), while none of the isolates was resistant to any of the 15 antibiotics. Both resistant and sensitive strains responded approximately equally to BC after adaptation, and remained stable during subculturing in the absence of BC. They suggested that genes coding for the efflux pumps providing resistance against QAC and ethidium bromide are not located on the multiple drug resistance (MDR) plasmid. When sublethal levels of a triclosan-containing domestic detergent were applied to a biofilm, the composition of the biofilm changed; howev