In the autumn of 1996, Laurie Girand, a marketing consultant in Silicon Valley, came home from a Fiji vacation to find her three-year-old daughter Anna in constant stomach pain. “My tummy hurts. My tummy hurts,” the usually resilient child kept saying. Within days, Anna was in the hospital, severely anemic, her red blood cells looking like shredded circles under the microscope. Doctors diagnosed her with hemolytic uremic syndrome, or HUS, a sometimes fatal complication of foodborne infection. The bacterium behind the disease, E. coli O157:H7, is usually linked to undercooked hamburgers—Girand vaguely remembered a huge Jack in the Box outbreak of a few years earlier. But her daughter hadn’t gotten sick from eating hamburgers. Only as Anna lay in a hospital bed, ashen-faced, glassy-eyed, waiting for a blood transfusion, did Girand and her husband hear a news account that apple juice had been linked to an E. coli epidemic. Manu-

factured by Odwalla, Inc., its selling point was that it was unpasteurized, and presumably more wholesome. The apple juice had been a treat from Anna’s grandmother while Girand and her husband were away. And though Girand didn’t approve of apple juice—too many empty calories—for a long time she had been feeding her daughter Odwalla unpasteurized carrot juice. “I was under the seriously mistaken impression that feeding our daughter unpasteurized juice would be healthier for her,” Girand says. “Odwalla’s slogan at the time was: ‘Drink it and thrive.’”

Though anemic for months, Anna recovered. But in Colorado, a child had died, while nearly 70 others, mostly children six and younger, had become severely ill in what was, until then, the country’s biggest juice-associated outbreak. Catalyzed by the near-tragedy, Laurie Girand started giving speeches to parents and writing letters to government officials. But shocking as the Odwalla outbreak was, it was not sufficiently instructive. Three years later, in the nation’s biggest juice outbreak, Salmonella in unpasteurized Sun Orchard orange juice struck nearly 500 victims and killed one. A few months later, the company had to recall another tainted lot. “I can’t believe,” Girand says, “since I am a marketing person, how badly I was fooled by industry marketing.”

One of the most insistent marketing messages we hear, trumpeted by both industry and regulators, is that the United States has the safest food supply in the world. Yet according to the CDC’s best calculations, each year 76 million Americans—nearly one in four, and that’s a lowball estimate—become infected by what they eat. Most find themselves for a few days dolefully memorizing a pattern of bathroom floor tiles. About 325,000 land in the hospital. Two million suffer drawnout, sometimes lifelong medical complications from unwittingly eating a contaminated morsel. More than 5,000—about 14 a day—die from indulging in what should be one of life’s great pleasures. The “world’s safest food supply” regularly doles out E. coli O157:H7 in hamburgers, Salmonella in alfalfa sprouts, Listeria in hot dogs, Campylobacter in Thanksgiving turkeys.

Change is what ushers new disease-causing organisms into our lives. And in the past few decades, there have been profound shifts in what we eat, where our food comes from, how it’s made, and who makes it. Fifty years ago, grocery stores stocked about 200 items, 70 percent of which were grown, produced, or processed within a 100mile radius of the store. Today, the average supermarket carries nearly 50,000 food items, some stores as many as 70,000. Agriculture and food manufacture have grown into global economies of scale, producing megaton quantities that, if contaminated, increase the potential for widespread epidemics. More fresh fruits and vegetables come from abroad, where sanitary standards may not be as high as in the United States. And our meals are increasingly cooked by people untrained in the techniques of safe food preparation.

This is not your grandparents’ “food poisoning”—a now-quaint term that originated early in the twentieth century, when dramatic gastrointestinal distress was usually traced to toxins, especially staph toxins, that had grown on spoiled foods such as cream-filled pastries or chicken salads left out too long in summer heat. Literally cases of food “intoxication,” these infections struck suddenly and fiercely, usually within two to six hours after the meal. When local health officials worked up these classic “point source outbreaks,” they would inevitably find that a knot of victims had all eaten a single dish, and that cases sharply climbed and then plummeted as the well of exposed individuals dried up. Point source outbreaks haven’t faded away; big-city health departments face dozens every year. In 1997, for instance, Salmonellatainted hams from a church fundraising dinner in St. Mary’s County, Maryland, sickened 700 people and killed an elderly woman. Today, however, the modest church picnic has given way to a giant food bazaar created by massive consolidation and global distribution. One contaminated tidbit—a shred of meat from an infected steer mixed with hundreds of other carcasses for hamburger, an iced box of tainted lettuce dripping down on the rest of an outbound lot, a soiled production line of cereal shipped coast-to-coast under 30 different brand names—spreads disease far and wide.

The pathogens in science’s crosshairs have also changed—in part

because improved technology permits scientists to see some for the first time, and in part because evolution has selected for more noxious creatures. Twenty years ago, today’s most fearsome threats were overlooked or yet-to-be-discovered. Campylobacter jejuni, now known to be the most common bacterial agent in food, was considered a rare, opportunistic organism because lab workers didn’t see it hiding among less fastidious bacteria growing in culture. A small, delicate, spiralshaped microbe, it corkscrews its way into mucous membranes of the intestinal tract “with a speed that cannot be matched by other bacteria,” according to one scientist’s report. Listeria monocytogenes, the most deadly agent in our food supply, killing one in five victims it infects, wasn’t even suspected of spreading through food. E. coli O157:H7, a potent threat to children and the aged, was identified only in 1982—and even then remained a medical curiosity until the infamous 1993 Jack in the Box hamburger outbreak. Norwalk virus, the top cause of foodborne illness in this country at 23 million cases a year, remained largely elusive until molecular tests revealed it in the 1990s. All of which suggests there are novel disease-causing agents still hiding incognito in our food. Even with modern diagnostic tools, in 81 percent of foodborne illnesses and 64 percent of deaths, doctors don’t know what organisms to blame—in part because they don’t know what organisms to look for.

To doctors and scientists, some of these bugs—particularly E. coli O157:H7—are scarier than anything seen before. “Foodborne pathogens are not purely a bit of nausea and vomiting and diarrhea,” says David Acheson, an E. coli researcher at Tufts University School of Medicine. “They can kill a previously healthy person in the space of a week.” Evolutionary biologists fear that our efforts to eliminate pathogens on the farm and in processing—by, for example, using disinfectant rinses—may paradoxically help select for more durable and virulent strains.

Meanwhile, more of us are more vulnerable to foodborne microbes. Individuals with impaired immunity—the very young, the very old, and people with cancer, organ transplants, diabetes, AIDS, and other conditions that weaken the body’s defenses; all told, about a

quarter of the population—are more apt to succumb to these infections. Men and women over 65, who in the next three decades will make up one-fifth of the population, produce less acid in their stomachs, eliminating the first line of defense against enteric pathogens; federal officials predict that the aging population could increase foodborne illness by 10 percent in the next decade. Americans are popping more prescription and over-the-counter antacids than ever, and in so doing, giving pathogens entrée to the nether regions of our digestive system where they do the most damage.

Depending on the organism, the palette of symptoms associated with foodborne disease can include diarrhea, cramps, fever, nausea, and vomiting (the notable exception is Listeria, which can cause miscarriage, meningitis, and other nonabdominal problems). But that’s just the beginning. In some people, researchers have discovered, the gastrointestinal distress that comes and goes with a foul meal may hang around in another form much longer. Salmonella can trigger reactive arthritis, an acute joint inflammation. Campylobacter jejuni may cause as many as 40 percent of cases of Guillain-Barré syndrome, a severe neurological disorder that can bring temporary paralysis and long-term nerve damage. Other complications include thyroid disease, inflammatory bowel disease, and, should someone survive the struggle against E. coli O157:H7, permanent kidney damage from hemolytic uremic syndrome. In these cases, contaminated food seems to provoke an uncontrolled autoimmune reaction. Up to 3 percent of foodborne disease victims—an enormous number, given the total caseload—may suffer lifelong physical problems.

Any depiction of emerging foodborne infections is necessarily panoramic, complex, and accompanied by more questions than answers. This discussion is no exception. As you will discover, debates about questions of farm management, government regulation, and individual versus institutional responsibility may elicit two—or three or four— diametrically opposed arguments that all seem persuasive. “Foodborne illness is more complex than people understand. The more I learn, the less I realize I ever knew,” says Mike Osterholm, a former Minnesota state epidemiologist who has probably launched more successful food

outbreak investigations than any other public health official in history. “The very nature of the ever-growing and complex food supply chain, and the desire of consumers to have many different kinds of foods available at a moment’s notice, has allowed for a whole new spectrum of pathogens to arrive on the scene.” What’s more, says Osterholm, DNA fingerprinting has pulled back the covers from foodborne out-breaks, showing that “many of the old conclusions we had drawn about what was happening are not valid.”

For those of us who don’t think about it for a living, it’s easy to underestimate the risk of falling ill from food, since the problem is largely invisible—hidden, one supposes, behind the bathroom door. CDC epidemiologists have factored in this cultural aversion by using numerical multipliers that translate the relatively few cases reported into a far higher and more accurate count of victims who never see a doctor. For instance, for every person known to suffer an infection caused by Campylobacter or Salmonella or Cyclospora, there are 38 who have eluded the net of public health officials; for every confirmed case of E. coli O157:H7, there are 13 to 27 doubled-over victims. Keep this in mind when you read news stories about foodborne epidemics. The scores of confirmed cases mentioned in wire service stories may actually represent hundreds or thousands of silent sufferers.

Foodborne infections are ubiquitous, sneaky, and regularly sold short. At the CDC, the foodborne and diarrheal diseases branch investigates more outbreaks than any other group in the agency. According to Paul Mead, a medical epidemiologist, “The paradox of foodborne illness is that, on a per meal basis, it’s extremely rare. It’s like getting hit by a meteor.” But in the very act of eating, says Mead, “You’re standing in a meteor shower three times a day from the time you’re weaned until you die.”

Every pathogen has a story, but the biography of E. coli O157:H7 is especially instructive because it shows how chance favors the prepared germ—and how we are giving certain disease-causing organisms

more chances than a rigged roulette wheel. Though E. coli O157:H7 has turned up in unpasteurized apple cider in 1991, 1996, and nearly every year since the Odwalla outbreak, it is best known as the agent behind “hamburger disease.” Hamburgers, in fact, are Rolls-Royce conveyances for O157. Think of your next Big Mac as the end product of a vast on-the-hoof assembly line. The story begins on hundreds of feedlots in different states and foreign countries. The animals are shuttled to slaughterhouses, where they become carcasses. The carcasses go to plants that separate meat from bone. The boning plants ship giant bins of meat to hamburger-making plants. The hamburger-making plants combine meat from many different bins to make raw hamburgers. At this point, your burger is more fluid than solid, because ground beef continually mixes and flows as it’s made, its original ingredients indistinguishable. Grinding also multiplies surface area, so that the meat becomes a kind of soup or lab medium for bacteria. Finally, from the hamburger-making plants, these mongrel patties are frozen and sent to restaurants. A single patty may mingle the meat of a hundred different animals from four different countries. Or, looked at from another perspective, a single contaminated carcass shredded for hamburger can pollute eight tons of finished ground beef. Finding the source of contamination becomes impossibly daunting. (Making juice is also like making hamburgers: one bad apple can ruin a huge batch.) In the Jack in the Box outbreak, investigators found that the ground beef from the most likely supplier contained meat from 443 different cattle that had come from farms and auction in six states via five slaughterhouses. As the meat industry consolidates and the size of ground beef lots grows, a single carcass may have even more deadly potential. In 1997, Hudson Foods was forced to recall 25 million pounds of ground beef for this very reason: a small part of one day’s contaminated beef lot was mistakenly mixed with the next day’s, vastly spreading the risk.

E. coli O157:H7, the organism that this endless mixing amplifies, is a quiet tenant in the intestines of the 50 percent or so of feedlot cattle it infects, but a vicious hooligan in the human gut. In the bowel, Escherichia coli, rod-shaped bacteria first described by German pedia-

trician Theodore Escherich in 1885, perform a vital task by keeping disease-causing bacteria from taking over. For many decades, that knowledge obscured the fact that some forms of E. coli trigger violent disease. E. coli O157:H7 (the letters and numbers refer to immune system–provoking antigens on the body and on the whiplike flagella of the organism) was discovered in 1982, during an epidemic spread by undercooked patties from McDonald’s restaurants in Oregon and Michigan. The outbreak wasn’t highly publicized; even some scientists perceived O157 as more of an academic curiosity than a harbinger of bad things. Eleven years later, the Jack in the Box hamburger chain promoted its “Monster Burgers” with the tag line: “So good it’s scary.” These large, too-lightly-grilled patties killed four children and sickened more than 700 people—bringing the exotic-sounding bacterium out of the lab and into public consciousness. In fact, however, by the time of the Jack in the Box tragedy, 22 outbreaks of E. coli O157:H7, killing 35 people, had already been documented in the United States. Suddenly, fast food hamburgers—a staple of American culture—were potentially lethal.

What makes E. coli O157:H7 so fearsome is the poison it churns out—the third most deadly bacterial toxin, after those causing tetanus and botulism. Known as a Shiga toxin, because it is virtually identical to the toxin produced by Shigella dysenteriae type 1, it is a major killer in developing nations. The distinctive symptoms of E. coli O157:H7 are bloody diarrhea and fierce abdominal cramps; many victims say it’s the worst pain they ever suffered, comparing it to a hot poker searing their insides. Between 2 and 7 percent of patients— mostly young children and the elderly— develop hemolytic uremic syndrome, which can lead to death. HUS sets in

when Shiga toxins ravage the cells lining the intestines. The bleeding that ensues permits the toxins to stream into the circulatory system, setting up a cascade of damage similar to that of rattlesnake venom. The toxins tear apart red blood cells and platelets, leaving the victim vulnerable to brain hemorrhaging and uncontrolled bleeding. Clots form in the bloodstream, blocking the tiny blood vessels around the kidneys, the middle layer of the heart, and the brain. As the kidneys give out, the body swells with excess waste fluids. Complications ripple through all major organ systems, leading to strokes, blindness, epilepsy, paralysis, and heart failure. Though doctors can manage HUS symptoms, and are working on new ways to stymie the toxin, they currently can offer no cure or even effective treatment.

For public health officials, the emergence of E. coli O157:H7 is an object lesson in how a new pathogen can lie low in the environment, biding its time until humankind changes a certain activity and in so doing rolls out a red carpet. Like other emerging pathogens, such as the AIDS virus, O157 had struck long before it caught the attention of public health officials. In 1955, a Swiss pediatrician in a dairy farm area first described HUS, which physicians today consider to be a gauge of E. coli O157:H7 infection. Over the ensuing years, the number of cases kept rising, suggesting that O157 was quietly spreading. In 1975, doctors took a stool sample from a middle-aged California woman with bloody diarrhea, cultured the apparently rare bacterium and sent it to the CDC, where it sat in storage until the McDonald’s outbreak prompted researchers to scour their records for earlier evidence of the vicious organism. In other words, for nearly 30 years before the first bona fide epidemic, E. coli O157:H7 had turned up in scattered, sporadic cases of bloody diarrhea. It was out in the meat supply, but not in high enough concentrations to catch health officials’ notice.

Where did E. coli O157:H7 come from in the first place? Scientists have pieced together a long, rather provocative history. Genetic lineages suggest that about 50,000 years ago, O157 and another closely related serotype—O55:H7, which causes infant diarrhea in developing nations—split off from the same mother cell. Since then, O157 has

taken part in a series of biological mergers and acquisitions that left it as vigorous as one of today’s giant pharmaceutical houses. Indeed, a 2001 study showed that O157, composed of more than 5,400 genes, picks up foreign DNA at a much faster rate than do other organisms. At some point, it acquired two deadly Shiga toxin genes after being infected by a bacteriophage, a tiny virus that insinuates its DNA into the chromosome of a bacterium. In the microbial world, phages are like squatters in Amsterdam, casually taking up residence in new bacteria, perhaps as a response to environmental stresses such as ultraviolet light or toxic chemicals. Bacteriophages are also the villains behind some of the most deadly human plagues; the genes coding for the cholera toxin, for instance, were borne on a phage. So what surrounding pressures compelled the phage carrying the Shiga toxin genes to light out for a new home in E. coli? In experiments on mice, Tufts University researcher David Acheson may have found the answer. When Acheson gave the animals low levels of antibiotics, the phage virus wildly replicated itself, and its magnified forces were more likely to infect other bacteria. Antibiotics also spurred the phage to pour out clouds of Shiga toxin. Acheson speculates that when farmers began the practice of feeding cattle small doses of antibiotics to spur growth, beginning in the 1950s—perhaps not coincidentally, when the first reports of sporadic HUS in children came out—they may have unleashed O157. More backing for this theory comes from epidemiological evidence. E. coli O157:H7 is a disease of affluent, developed nations— which also happen to be the ones that feed growth-promoting antibiotics to livestock.

What worries Acheson and other scientists is that the restless phages that manufacture Shiga toxin may jump to other disease-causing bacteria. Actually, they’ve already proven they’re disposed to do this, having set up home in about 200 other strains of E. coli. One of these, E. coli O111:H8, in 1999 caused a massive epidemic of nausea, vomiting, bloody diarrhea, and severe stomach cramps at a high school drill team camp in Texas, sickening dozens of the 750 teenage girls who attended. Though investigators never did find where the organism was hiding, they suspect it was either in the ice the girls used to

soothe their parched throats during the drills or somewhere in the salad bar. Shiga toxin phages have also landed in Enterobacter and Citrobacter—other bacteria that stir up intestinal disease. To find out just how prevalent these mysterious strains of dangerous E. coli may be, Acheson analyzed ground beef samples from 12 supermarkets in Boston and Cincinnati. The results came as a shock. He found Shiga toxin in a quarter of the samples—toxin produced not by O157:H7, but by other kinds of E. coli. And this may not be the end of their roving, Acheson warns. “Suppose something like Salmonella developed the ability to produce Shiga toxins. That could be an extremely deadly pathogen.” Not only is Salmonella common, but, more than E. coli O157, it has a talent for quickly invading the bloodstream, meaning it could speedily convey Shiga toxins throughout the body like tiny poison-tipped missiles. Even more problematic, the antibiotics normally used to treat E. coli O157:H7 infections may actually aggravate the illness, by kicking phages into overdrive and stepping up their production of toxins, leading to hemolytic uremic syndrome.

Along its evolutionary path, E. coli also became acid resistant, so impervious to a low pH environment that it can survive the incredibly sour bath in the human stomach. Grain-feeding cattle, which supplanted traditional hay feeding after World War II, may have made the bacteria more acid resilient. Because of this acid tolerance, as few as 10 organisms are enough to cause infection. Having acquired a mean set of toxin genes, acid resistance, and other virulence properties, all E. coli O157:H7 needed to become a truly fearsome threat was access. That it acquired by spreading in domesticated cattle and then entering the gears of modern industrial meat production, all within the past 25 years. Unfortunately, O157 may have left the door open behind it. Other strains of E. coli, “if tweaked in the right way” by phages and the mobile rings of DNA known as plasmids, could negotiate the same path, says Tom Whittam, a biologist at Pennsylvania State University who has studied O157 evolution.

Research is under way on vaccines that would prevent cattle from carrying O157, and on feed additives—including competing intestinal bacteria—that would eliminate the pathogenic organism in livestock.

Thoroughly cooking ground beef to a temperature of 160 degrees Fahrenheit is the proven method of killing E. coli O157:H7. But in the United States, the organism retains a fighting chance because of the American love affair with rare burgers, which practically guarantees that one man’s meat will be another man’s poison. As a restaurant menu in suburban Dallas proudly informs its customers: “The Department of Health suggests MEDIUM-WELL for any ground beef product. Our burgers are cooked MEDIUM (PINK) unless you request otherwise.”

The E. coli O157:H7 saga shows how the denizens of an animal’s GI tract find their way to our own digestive systems. This brings up a delicate point, rarely discussed in polite company, but one central to the rest of this chapter. Put simply, animal and human waste is the source of most foodborne illness. And what we eat usually becomes contaminated long before it reaches us—during processing, at the slaughterhouse, or right on the farm.

Of course, that’s a resonant theme in public health. The sanitary revolution of the nineteenth century—the discovery that the diseases of squalor and overcrowding could be prevented with sewage removal and clean water—was occasioned by fear of cholera, typhoid fever, and other pestilential diseases. Before this transformative event, daily life was unimaginably filthy. “Thousands of tons of midden filth filled the receptacles, scores of tons lay strewn about where the receptacles would receive no more,” observed an English medical officer in Leeds in 1866. “Hundreds of people, long unable to use the privy because of the rising heap, were depositing on the floors.”

Which is precisely how the animals that become our food live today. And why, at the CDC, officials in the foodborne and diarrheal disease branch long for a sanitary revolution: clean piped water and sewage disposal and treatment—for animals. Like the nineteenth century innovations that controlled typhoid fever, animal sewage would be separated from the human food and water supply, and also from the

animal food and water supply. “It’s a paradigm shift,” says the CDC’s Fred Angulo. “Farmers don’t consider themselves food handlers.”

The site of modern meat production is akin to a walled medieval city, where waste is tossed out the window, sewage runs down the street, and feed and drinking water are routinely contaminated by fecal material. Each day, a feedlot steer deposits 50 pounds of manure, as the animals crowd atop dark mountains composed of their own feces. “Animals are living in medieval conditions and we’re living in the twenty-first century,” says Robert Tauxe, chief of the CDC’s foodborne and diarrheal diseases branch. “Consumers have to be aware that even though they bought their food from a lovely modern deli bar or salad bar, it started out in the sixteen hundreds.”

The feedlot is just the start of their fetid journey. At the head of the slaughterhouse line, a “knocker” wields a pistol-like device to drive a metal bolt into a steer’s head. Other workers cut the animal’s throat to drain blood, and use machines to sever the animal’s limbs, tear off its hide, pull out its organs. More than 300 animals may pass through the line in an hour, each carcass weighing 650 to 800 pounds. At the slaughterhouse, writes journalist Eric Schlosser, “The hides are now removed by machine; but if a hide has not been adequately cleaned first, pieces of dirt and manure may fall from it onto the meat. Stomachs and intestines are still pulled out of cattle by hand; if the job is not performed carefully, the contents of the digestive system”—i.e., waste—“may spill everywhere.”

A United States Department of Agriculture study published in 2000 found that 50 percent of feedlot cattle being fattened for slaughter during the summer months carried the E. coli O157:H7 bacterium in their intestines—a far higher figure than previous government estimates. Another study found that about 43 percent of the skinned carcasses tested positive before being eviscerated, suggesting that microbes were being spewed within the plant.

In early July 2000, the Excel Corporation—the nation’s second-largest beef processor—allowed an Associated Press reporter to visit its huge Fort Morgan, Colorado, meat packing plant. Asked about the dangers of tainted meat reaching consumers, Excel’s food safety direc-

tor replied: “It’s like a roll of the dice or a game of Russian roulette.” Two weeks later, the face of three-year-old Brianna Kriefall, of South Milwaukee, appeared on front pages across the country. She had died from eating a slice of watermelon at a Sizzler restaurant. The watermelon had been sliced in the restaurant kitchen, on the same countertop where a meat grinder was used to convert steak trimmings—E. coli–contaminated steak trimmings—into hamburger. The trimmings came from sirloin meat packed in heavy vacuum-sealed bags. The bags had been shipped just a few days earlier from Excel’s Fort Morgan plant.

***

Chicken farming is just as noxious. But before delving into that, a word about chickens: they’re not all created equal. In the agribusiness world, there are two kinds of chickens—the broilers that give us meat, and the layers that give us eggs—and they are totally separate industries governed by different practices, riddled with different problems, and even centered in different parts of the country (the top broiler states are Georgia and Arkansas, while the top egg-producing locales are Ohio and California).

First, a look at broilers. In her book Spoiled, journalist Nicols Fox writes that “If chicken were tap water, the supply would be cut off.” Oddly enough, the government doesn’t have hard numbers on Salmonella and Campylobacter contamination rates, and what they do have is hardly appealing. A 1999 study from the USDA’s Agricultural Research Service, for instance, found that 7 percent of chickens sampled at slaughterhouses had Salmonella and 30 percent had Campylobacter— but, as one scientist there admitted, those numbers are probably low. For many years, researchers assumed that clearing feces, rodents, and insects from the broiler houses where the birds live out their five to nine weeks would solve the problem. But new studies suggest that the source of chicken contamination may be more deep-rooted. The 9.5 billion young broilers that Americans eat each year are actually the fourth generation in a carefully husbanded line. Scientists now believe

it’s the three previous generations—the “breeders”—that regularly pass down infection. When he tested birds at the top of the pyramid —the great-grandparent breeder flocks—microbiologist Nelson Cox at the USDA’s Russell Research Center found that 36 percent were positive for Salmonella. Cox suspects that these birds transmit pathogens to subsequent generations by contaminating their own eggs with feces that carry high levels of Salmonella, Campylobacter, Listeria, and Clostridium perfringens. Because the hen’s body temperature is quite warm—between 104 and 107 degrees Fahrenheit—she usually lays her egg on a day in which the air is cooler than her body. That temperature gap forces bacteria on the porous surface of the egg to get sucked into the membrane underneath, where most organisms live contentedly while the fertile egg is incubating. When the chick pecks its way out, it eats the pathogens. “The largest contributor to contamination of a broiler flock,” says Cox, “is the mother hen—the feces of the parent bird.” That means the human disease on our end of the food chain won’t end until farmers either clean up the three generations above, or scientists figure out how to snap the links of contamination. Vaccines may not be the answer, since they are only effective against diseases that make chickens sick—and both Salmonella and Campylobacter are benign commensals, living happily in the birds’ intestinal tracts without causing harm. Another possibility, slaughtering the priceless greatgrandparent breeder birds, would drastically raise chicken prices.

Now on to layers. Modern houses for egg production are avian megalopolises. In 1945 the typical henhouse sheltered 500 birds; today it can contain 80,000 to 175,000, with up to 20 houses in a single operation. (As in the livestock industry, this huge scale is a result of industry consolidation; in 1996 there were approximately 900 egg operations in the United States, compared to 10,000 in 1975.) Laying flocks stay in the same house for up to a year and a half, which means that detritus builds up. “You have a lot of everything,” says Richard Gast, a microbiologist at the USDA’s Southeast Poultry Research Laboratory. “A lot of birds, a lot of manure, a lot of moisture, a lot of dust. Everything that walked into that house—every two- and four- and six-legged creature—is a potential vector for moving it around.”

What spreads in this tenement is Salmonella enteritidis, or SE, the villain behind most egg-related outbreaks. SE is a versatile bug, capable of infecting birds through two different routes. One is orally, since chickens eat feces. Another route—more troubling, because scientists haven’t figured out how to interfere with it—ascends through the cloaca, the cavity in birds into which empties the products of both the intestinal and reproductive organs. SE is sucked up into the bird’s reproductive tract and eventually into the ovaries. From there, it gets inside eggs even before the shell is laid down—indeed, most eggs become systemically contaminated with Salmonella enteritidis in this way. Just where SE came from, or why it spread so suddenly in the 1980s, remains a mystery. Found in 1 of every 20,000 eggs, SE makes French toast, Hollandaise sauce, and raw cookie dough risky culinary excursions.

Animal waste and its dangerous microbes aren’t confined to the farm, of course. Manure—spread through fertilizer, irrigation water, insecticide solutions, dust, even wild birds and amphibians—gets on produce too. Typical is an outbreak that took place in 1998, when patrons of a Kentucky Fried Chicken restaurant in Indianapolis became ill with E. coli O157:H7. Zeroing in on KFC’s cole slaw, investigators discovered that some of the cabbage came from fields supplying a Texas vegetable company—and that, during a severe drought, the fields were flooded with untreated water from the Rio Grande, where cattle had waded and relieved themselves in the irrigation canals. Similarly, at Disney World in Orlando, Florida, thousands of visitors from all over the country were believed to have been infected with Salmonella in 1995 after drinking unpasteurized orange juice at special “character breakfasts” at the park, in which costumed Disney characters mingle with the guests. The orange juice came from a small processing plant nearby—a plant where the walls and ceiling of the processing room had cracks and holes, and where frogs congregated near the equipment. Outside the plant, investigators found Salmonella in a toad, in tree frogs, in soil, and on unwashed oranges. E. coli O157:H7 contaminates unpasteurized cider when fallen apples touch cattle or deer waste and are then mixed with other pieces of fruit. Numerous lettuce

outbreaks have occurred after the heads were exposed to cattle manure. Organic foods are hardly immune to these pitfalls. In fact, microbiologists have found more bacterial contamination on organic than on conventionally grown produce, and no one is quite sure how much composting it takes to knock off pathogens in manure. That gap in knowledge has real-life consequences. In rural Maine in 1992, a woman who abided by a lacto-vegetarian diet, consisting almost exclusively of vegetables fertilized with manure from her cow and calf, developed E. coli O157:H7 when she failed to wash the vegetables well enough. Through improper handwashing, she passed the infection on to three neighborhood children, one of whom, a three-year-old boy, died of kidney failure.

Farm conditions create a wide-open channel down which emerging pathogens travel from food animals and produce to people, and the modern food industry has converted a two-lane country road into a 12-lane interstate. “Salmonellosis is rare in developing countries, where sanitation is poor and diarrheal diseases are endemic, but where food production and consumption are local,” writes Martin Blaser, chairman of the department of medicine at New York University, in the New England Journal of Medicine. Blaser’s dispiriting conclusion? “Salmonellosis—with the notable exception of typhoid fever—is a disease of civilization.”

And outbreaks are not so much “point source” as pointillist. Changes in agriculture and food manufacture—vaster and fewer farms, slaughter plants, and processing facilities—have given pathogens a larger stage on which to strut. In this miraculous food economy of scale, when things go wrong, they go wrong in a big way. Mass-distributed items with spotty or low-level contamination are consumed by people living far from the source. This leads to a new, insidious kind of epidemic: one with low attack rates (less than 5 percent of the people who eat the contaminated food) but huge numbers of dispersed victims. Take the massive 1994 outbreak of Salmonella enteritidis. Usually, SE is linked to undercooked eggs or egg products. But Schwan’s ice cream, made in Minnesota and delivered to homes in all 48 contiguous states, was made from premix that had been transported to the

plant in tanker trailers—trailers that had previously carried unpasteurized liquid eggs. Though the insides of the tankers were supposed to be washed and sanitized after hauling eggs, drivers sometimes skipped that laborious step. Across the country, an estimated 224,000 ice cream aficionados—mostly kids—paid the price in the largest outbreak of salmonellosis ever recorded from a single food source.

The antibiotics that food animals eat can also make you sick. Though the drugs are used to fatten livestock and protect them from disease, they have the paradoxical effect on humans of breeding mean, antibiotic-resistant infections. Here’s how the process works: In an animal’s gut, antibiotics foster the growth of bacteria such as Salmonella, Campylobacter, or E. coli that resist the antibiotics. If you eat undercooked meat from that animal, you swallow those antibiotic-resistant bacteria, and you may or may not get sick. By themselves, drug-resistant organisms in food don’t necessarily trigger symptoms, because the bugs are held in check by other bacteria in the gut. But if you happen to be taking the antibiotic to which the organism is resistant— say, tetracycline—you can get very sick. That’s because the drug clears out other benign bacteria in your intestines, opening the way for the very pathogen that resists the antibiotic to run rampant in your colon and sometimes beyond.

“The reason we’re seeing an increase in antibiotic resistance in foodborne diseases is because of antibiotic use on the farm,” says the CDC’s Fred Angulo. In the United States, an estimated 70 percent of the antibiotics produced each year—nearly 25 million pounds, according to a 2001 report—goes to food animals, in low nontherapeutic doses. Farmers mix these antibiotics in animal feed for two reasons. One is to prevent disease. The other is to promote growth and boost the conversion efficiency of feed into flesh, though exactly how low-dose antibiotics accomplish this isn’t clear. It may be that these drugs kill off not only disease-causing bacteria in the gut, but also the good bacteria that compete for nutrients. What’s scary is the overlapping of

farm and pharmacy. Of the 19 classes of antibiotics used in animals as growth promoters, seven are prescribed for people.

When foodborne pathogens turn antibiotic-resistant, they wreak all kinds of havoc. They are more virulent, and they afflict more people because it takes fewer organisms to cause infection. Patients with antibiotic-resistant infections stay in the hospital longer. The infections especially threaten children, the elderly, and people whose immune systems are weak, such as cancer or AIDS patients: all groups likely to take antibiotics. Foodborne infections that breach the intestinal tract and enter the rest of the body trigger bloodstream or nervous system infections—for which antibiotic treatment can be lifesaving. When the pathogen resists the best drugs doctors can offer, death rates climb. Resistant foodborne infections also complicate treatment for other, unrelated infections.

For many years, farmers and regulators were resolutely skeptical that antibiotics in animals could have downstream effects in people. It took a dramatic 1983 outbreak, in which 18 people in four Midwest states came down with a ferocious strain of antibiotic-resistant Salmonella newport, to erase the conventional wisdom. Just before becoming ill, most of the patients happened to have taken a form of penicillin for garden-variety infections: bronchitis, earaches, strep throat. So dramatic was the link between taking the antibiotic and coming down with salmonellosis—patients were 51 times more likely than those in the control group to have taken the drug—that public health officials first suspected the antibiotic itself was contaminated. The truth was much more devious. When investigators gathered patients’ food histories, they found that all had eaten ground beef shortly before falling ill. In each case, the hamburger meat had come from a South Dakota farm where beef cattle were fed “subtherapeutic” doses of antibiotics. On the adjacent farm, a dairy calf had died of Salmonella newport. Investigators conjectured that the dairy herd somehow transmitted the bacterium to the beef cattle which, being fed small doses of tetracycline antibiotics, went on to develop resistant strains of the organism. The infected beef cattle contaminated at least 40,000 pounds of ground

beef with antibiotic-resistant Salmonella; without a doubt, many more people ate it than came to health officials’ attention.

Today, the link between antibiotic use on farms and human disease is richly documented. In Muslim countries, resistant foodborne bacteria in people are almost always identical to microbes found in poultry—not surprising, since pork is banned. In the Netherlands, vancomycin-resistant enterococci, or VRE—a looming problem in hospitals—are noticeably absent from the intestines of strict vegetarians. Around the world, governments have been embroiled in the question of whether food animals should get the same antibiotics prescribed for people. This question is bound to loom larger, since—as you will learn in the next chapter—many public health officials find antibiotic-resistant infections to be the most terrifying prospect on the horizon.

Another insidious foodborne infection may be looming on the U.S. horizon. New variant Creuzfeldt-Jakob disease—the human form of mad cow disease—has proceeded narrowly and stealthily through a food chain whose links are masked by intensive food production and globalization. At this writing, more than 100 cases, invariably fatal, have been reported, mostly in Britain with a handful in Europe. What researchers don’t know is whether these represent the waning aftermath of a narrowly averted public health disaster or the first rumbles of a terrible storm.

“Mad cow disease” is a term that had not even been coined in late 1984, when a veterinarian called to a farm in West Sussex, in southern England, found a dairy cow displaying “a variety of unusual clinical manifestations”: panic, aggression, a staggering gait. By the spring of 1985, more cows came down with the mysterious malady, later named bovine spongiform encephalopathy, or BSE, the giveaway marker of which was brain tissue that resembled Swiss cheese. By the late 1980s, scientists began piecing together the puzzle. The epidemic likely began as a foodborne outbreak among livestock. In the early 1980s, cattle

were fed remnants from sheep infected with scrapie, a brain-wasting disease (discarded animal parts are considered a cheap source of protein that increases milk production). In a further perversion of nature, the inedible parts of these infected cattle were themselves made into meat and bone meal for other cattle, a thrifty practice that not only permitted the recycled infectious agent to amplify but also to adapt to its new host. Changes in the rendering process also helped the agent survive. Before 1981, the carcasses of ruminants had been subject to high heat and organic solvents to remove fat and disarm disease-causing proteins, such as viruses. But that year, various economic factors persuaded manufacturers to turn down the heat and cut out the solvents, allowing the yet-to-be-discovered infectious agent to escape inactivation. Though hundreds of thousand of cows would eventually be diagnosed with BSE, and millions of animals destroyed as a precaution, the British government was stalwartly optimistic that the epidemic would stay put on the farm. As the Southwood Report noted in 1989, “It is most unlikely that BSE will have any implications for human health.” By 1996, after a new variant of the neurological affliction Creutzfeldt-Jakob disease—featuring bizarre behavioral and personality changes, staggering, and dementia—appeared in startlingly young patients, the scientists changed their tune, to the horror of a carnivorous nation. “Beef is one of the great unifying symbols of our culture,” lamented a Guardian editorial. “The Roast Beef of Old England is a fetish, a household god, which has suddenly been revealed as a Trojan horse for our destruction.”

How much beef contaminated by prions—abnormally folded proteins—do humans have to consume to become infected? No one knows. And no one knows how long the incubation period is for new variant Creutzfeldt-Jakob disease, though judging from events in Britain it seems to be at least 10 to 15 years. No one knows how much BSE-infected beef was slaughtered for human consumption before the epidemiologic puzzle was pieced together—perhaps 750,000 animals, perhaps a million; the UN estimated that at the height of the mad cow epidemic, Britain dumped 500,000 tons of untrackable bovine

byproducts on Western Europe and other nations. No one knows whether mad cow prions have infected people through blood or organ donations, contaminated surgical instruments, or consumer products and drugs that contain bovine material. The upshot of all this uncertainty is that no one knows where we are on the epidemiologic curve: at the end or the beginning of an outbreak? So murky is the science, Oxford University’s esteemed epidemiologist Roy Anderson calculated that human cases could conceivably range between 63 and 136,000, while a British government study put the high-end figure at 250,000.

Will the human variant of mad cow disease turn up in the United States? “The odds are that sooner or later we will see a case here,” says CDC director Jeffrey Koplan, “whether it’s an imported one, whether it’s home grown, whatever. None of us should be surprised if we have a case in the next week or the next ten years.” As of the fall of 2001, no cases have been reported. To prevent the spread of BSE to American farms, the U.S. government in 1989 banned the importation of live cows and sheep. Since then, it has erected regulatory fences to screen out other bovine products and has upgraded surveillance for brainriddling spongiform diseases in domestic animals and humans. The American Red Cross has tightened its blood donation rules for people who have been to Europe. Critics say these measures aren’t enough— that, to borrow from W. C. Fields, the United States has failed to “take the bull by the tail and face the situation.” The U.S. still imports biomedical products, for instance, that contain materials made from ruminants in countries harboring mad cow disease. In 2001, the Food and Drug Administration reported that companies were using ingredients from BSE countries to make nine widely used vaccines, including those for polio, diphtheria, and tetanus. The FDA has also failed to regulate dietary supplements such as those claimed to stimulate energy, sexual vitality, and memory—all of which can contain nervous system, organ, and glandular tissue from cattle. And the surveillance net for potentially infected cattle in the United States has big holes, while inspection of feed manufacturers and rendering companies is lax. Meanwhile, some scientists worry that spongiform disease could

strike Americans, not through the consumption of beef, but of hunted wild animals such as deer and elk, which are succumbing to another prion-related epidemic, chronic wasting disease.

In echoes of the mad cow crisis, another agricultural infection struck Britain in 2001—foot and mouth disease, one of the most contagious of all animal diseases. How did the culpable agent—in the same family as the common cold virus—get there? Likely from contaminated meat smuggled into Britain from countries where the disease is rife. The British army admitted supplying untreated waste food to a pig farm in Northumberland, where the virus incubated and then wafted over air currents to a flock of sheep. By the time the disease was identified days later, the virus had spread all over the country through markets and dealers. The government response was swift and shocking. Bonfires of livestock carcasses shot flames into the night sky—one writer described the giant pyres as “archaic precautions.” Europe and the United States, long protected against the infection, went on red alert, disinfecting the shoes of hundreds of thousands of arriving airline passengers from the British Isles—another reminder that the world is not just a global village, but a global pathosphere.

Consider the alfalfa sprout: from a moist mat of tendrils, a sinuous pale stem rises to a dark green, double-bladed capital. Sprout growers proudly call their crop a “living food,” in contrast to the butchered, disease-bearing livestock that are a staple of the American diet. They boast that one ounce of sprouts contains more protein than a pound of steak and is loaded with vitamin C. Prefiguring the counterculture, Captain James Cook raised lentil sprouts to stave off scurvy, until colonial economics made limes more favored on the high seas.

Food microbiologists hold a somewhat less romantic view of this ’60s holdover. “Sprouts are about as hazardous a food as you can get,” says Mike Doyle, director of the Center for Food Safety and Quality Enhancement at the University of Georgia. “You’ve got water, you’ve got the right temperature, and you’ve got nutrients. Those are the three

things you need for growing harmful microorganisms.” From a public health perspective, sprouts perfectly illustrate what can go awry when a “living” food that happens to be contaminated multiplies and goes forth into the food supply. And it shows that whenever a new food technology arrives on the scene, its impact on foodborne pathogens must be thoroughly pondered.

Between 1970 and 1999, the sprout industry grew at an annual clip of 10 percent. Many of its entrepreneurs were devoted vegetarians and organic gardeners who eschewed chemical additives such as pesticides, hormones, or disinfectants. By the end of the century, 15 different varieties of sprouts appeared in U.S. stores, from alfalfa and clover to more exotic species such as sunflower, wheatgrass, and anise. What these highly independent small businesspeople overlooked was that Salmonella and other pathogens are often present at the creation— lurking in the crevices of sprout seeds.

Described by Nicols Fox as “the cockroach of the microbial world,” the Salmonella genus is hardy, tenacious, adaptive, omnipresent, and virtually impossible to eradicate. Salmonella’s more than 2,400 serotypes can survive an amazing variety of conditions, from fish meal to fresh orange juice, from a hen’s ovaries to a turtle’s toes. Salmonella typhi causes typhoid fever, spread by contaminated water and food. Other Salmonella serotypes were named for the places where they were isolated, little flags on a global map of bad meals: S. newport, S. muenchen, S. heidelberg, and so on. Given its cosmopolitan ways, it’s no surprise that Salmonella turned up in sprout seeds. These seeds can be contaminated in the field from dirty water, runoff from nearby farms, and animal fertilizers, and from the feces of birds and rodents anywhere along the line from growth to storage to shipment. Salmonella organisms can survive for months under the dry conditions in which seeds are stored. The presence of even a few salmonellae on seeds is dangerous, because sprouting’s moist, warm growing conditions encourage bacteria to surge. In three to five days, ten organisms on a single seed can grow to more than 220 million in a typical eightounce container. You wouldn’t be able to see them with the naked eye or taste them in a salad—or wash them off, since they are found in the

tissue of the sprout, not just on the surface. And while the FDA recommends that sprout growers first rinse seeds with chlorinated water to bring down bacterial counts, the method doesn’t guarantee that pathogens will be eliminated before the crucial sprouting step. Making matters worse, sprouts are rarely cooked—a traditional kill step for pathogens—before being eaten.

The tipoff to the hazards of sprouts came in 1973, when a home sprouting kit spread Bacillus cereus, a toxin-forming organism found in soil. Since then, there’s been a steady stream of outbreaks, mostly from clover and alfalfa sprouts. In 1996 in Japan, white (daikon) radish sprouts infected an astounding 10,000 people with E. coli O157:H7. Just how rife Salmonella is became apparent in 1999, when the National Center for Food Safety Technology, a lab affiliated with the FDA, tested a new Salmonella rapid detection assay. Researchers inoculated Salmonella in some seeds and kept the control group of seeds clean. Perplexingly, the rinse water for both groups of seeds turned up positive for Salmonella. At first, the scientists thought the new test was giving false positive results—until they realized that the supposedly “clean” seeds in the control group not only were contaminated with Salmonella, but came from the same lot that was causing an outbreak in the Midwest.

Not until the spring of 1995 did U.S. public health officials wise up to the true dangers of sprouts. That year, the CDC kicked off a new method of discerning whether Salmonella cases were rising in this country. “SODA,” or Salmonella Outbreak Detection Algorithm, compared new state health department reports of Salmonella with the number that would be expected, based on historical data. The very next month the new system fulfilled its promise, revealing an international epidemic of Salmonella stanley that had been brewing for months, an outbreak that reached from Arizona and Michigan to Finland and probably Canada as well. Finnish officials who had already investigated the outbreak found that most of the patients remembered eating alfalfa sprouts before becoming ill. Tracing back the sprouts to their seed origins yielded a typically Byzantine course of events behind a modern international outbreak. The sprouts eaten by the American

patients came from at least nine different domestic growers, all of which had bought seeds from a single supplier to the U.S. market. This supplier had bought its seeds from various sources, including a shipper in the Netherlands. The Dutch shipper had mixed several lots of seeds purchased from Italy, Hungary, and Pakistan—and through an intermediate supplier, shipped some of these mixed lots to Finland. Ultimately, investigators estimated, between 5,000 and 24,000 people became sick in this low-profile, globe-spanning outbreak.

As the sprout outbreak showed, national borders dissolve in the face of foodborne pathogens. The General Accounting Office estimated that in 1995, one-third of all fresh fruit consumed in the United States was imported, a testament to lower production costs abroad and rising consumer demand for year-round fruits and vegetables. Seasonally, more than 75 percent of fresh produce may be harvested beyond U.S. borders. Though most imported food is wholesome, it is inevitable that, as the global food market grows, so will the number of international outbreaks. The question is: whose food standards will reign, those of America or those of our aspiring trade partners?

In the summer of 1998, when Minnesota health officials were barraged with outbreaks of gastrointestinal illness, they weren’t thinking about these questions. Shigella sonnei, a diarrhea-causing bacterium, had struck hundreds of patrons from two Minneapolis restaurants that apparently had nothing in common. One served Up North cuisine in Bunyanesque portions, while the other was part of a suburban horse and hunt club. The timing couldn’t have been worse. The health department was already besieged with a mysterious epidemic of enterotoxigenic E. coli, or ETEC—commonly known as “traveler’s diarrhea”—that had sickened more than 50 at a Twin Cities restaurant

and a lake resort. It was strange, this gastrointestinal tempest. In a state reputed to have the most vigilant and progressive public health department in the country, where national outbreaks were solved before other states even had an inkling something was amiss, health officials were stumped.

The Shigella bacterium is so potent that its misery can be transferred in a handshake. As few as ten organisms, a remarkably low infectious dose, can cause infection. Bacteria invade the cells lining the colon, inflame the intestines, and eventually bring on fever, doublingover abdominal pain, nausea, vomiting, and diarrhea. In the mid-1990s, a strain of Shigella sonnei, the predominant species in the United States, spread through eight communities of traditionally observant Jews in North America, fanning out in informal day care, ritual handwashing, and steam baths—more than 1,000 cases in a closed but global neighborhood. Today, the organism is most notorious for racing through day care centers, where toddlers and young children, whose personal hygiene practices are less than rigorous, spread infection—often to their diaper-changing parents and other adult handlers, who in turn infect others.

In Minnesota’s 1998 outbreak, suspecting at first that restaurant workers were the source of infection, investigators embarked on their standard gumshoe procedure. They interviewed the sick patrons and their unaffected dinner companions, hoping to pinpoint what the shigellosis victims had eaten. No food seemed resoundingly guilty. More surprising were the results from the state lab’s genetic analysis of the bacterium found in patients’ stools: the patrons of both restaurants had the same strain of Shigella. And this was a brand-new strain, not one that had been circulating around town. That meant the disease wasn’t spread by a few infected workers, usually the first guess in cases like these, but rather by a common food. And this food, whatever it was, must have come from the same farm. The investigators recrunched their data. They broke down every menu item at each restaurant into its constituent ingredients, trying to find what single element in two different kitchen larders could have been the culprit. The item that finally jumped out of a mess of statistics was a stunner: parsley. Specifi-

cally, chopped fresh curly parsley, decorative rather than delectable, the culinary equivalent of wallpaper.

Minnesota health officials launched a national investigation, dispatching a picture of the Shigella’s DNA on PulseNet, an electronic server that the CDC and state public health labs use to compare the genetic profiles of foodborne bacteria sampled at different sites. PulseNet is a public health version of the FBI’s 10 Most Wanted List, except that it features thousands of microbial criminals, and the mug shots it distributes are not the haggard faces of human reprobates but a column of shadowy bands representing unique stretches of DNA. Minnesota’s PulseNet image was accompanied by a question: Were other states also apprehending this strain of Shigella sonnei?

In Los Angeles, alarms went off. The city had seen two perplexing outbreaks of Shigella sonnei, one at an upscale steak-and-pasta place, the other at a Middle Eastern restaurant. Prompted by Minnesota’s message, investigators returned to the evidence and found that the strains from their two outbreaks not only matched each other, but also matched Minnesota’s—and that sure enough, the sick customers had all eaten parsley, either as a garnish or mixed into tabouli and falafel.

Soon Massachusetts chimed in with Shigella isolates that matched Minnesota’s. There had been an unsolved Shigella sonnei outbreak at a restaurant-and-bar joint north of Boston. The sick customers had all eaten grilled chicken breasts—which, investigators now learned, were liberally sprinkled with minced parsley.

A Marathon Key, Florida, restaurant was ground zero for Shigella sonnei. When investigators interviewed patients and other customers, they found that all had eaten dishes with fresh parsley—in fact, it was nearly impossible to find anyone who had not eaten parsley. The decorative sprigs were scattered over every entrée as if they had talismanic powers: on the baked mango snapper and fried grouper fingers, on the shrimp primavera and veal piccata, on the chicken cordon bleu and steak au poivre.

Canadian authorities also phoned: two outbreaks of Shigella sonnei, one in Alberta and one in Ontario, with the same strain as the one bombarding the States.

All told, 486 confirmed cases of shigellosis—and that was merely the count of diagnosed cases. In Minnesota alone, authorities estimated the disease actually struck at least 600 people, including visitors to the Goodhue County Fair. As it happened, several victims of the restaurant outbreaks, still suffering symptoms, had visited the fair and used the restrooms. The antiquated septic system leaked into the fair’s drinking water, extending a parsleyborne Shigella outbreak into a secondary waterborne epidemic (much like the events behind the massive 1999 E. coli O157:H7 outbreak at an upstate New York county fair that killed two and sickened more than a thousand).

So how had the parsley become contaminated in the first place? By Labor Day, Minnesota health officials had followed the product’s paper trail to a farm in Mexico. When state epidemiologist Mike Osterholm alerted the FDA, he was rebuffed. “I would have gotten more response out of talking to a rock. They said, ‘You don’t have absolute proof it’s there.’ I said, ‘I know we don’t have proof, in the sense of absolute. But our traceback takes it there.’” Not until that October did the CDC and FDA conduct an official “traceback,” a tedious and time-consuming physical examination of records from purchasers and assorted middlemen. The trail for seven of the eight outbreaks led to a vegetable growing operation south of Ensenada, Mexico, on the Baja Peninsula, about 100 miles south of the California border. Sprawled over 1,000 hilly acres hugging the Pacific, drenched in sunlight year-round, the farm grew tomatoes, celery, cilantro, green onions, radishes, and lettuce—a salad bowl inventory that would later add a twist to the story. Also under cultivation were about 55 acres of parsley—from a distance, brilliant emerald squares amid the brown hills. The herb was harvested early in the morning, chilled and packed in ice, and dispatched in a refrigerated tractor-trailer by 2 p.m.

During any traceback of contaminated produce, investigators are keenly curious about water. Water is used to irrigate crops, wash them, chill them, ice them. And water can come from many places—rivers, streams, irrigation ditches, wells, sewage treatment facilities, runoff, industrial washes, and ice. Water containing animal waste can inad-

vertently wash onto fields from swollen rivers or be mixed with pesticides. People suffering gastrointestinal illness can contaminate water or ice just by dipping their not-quite-clean fingers in it. When a CDC environmental engineer went to the Mexican farm, he checked out all these possibilities. He examined the irrigation system, fed by groundwater wells sunk in an old riverbed. He looked at the portable toilets in the fields, and found no toilet paper inside—to curb theft, supervisors required migrant workers to ask for it before they went to the bathroom. Yet while these conditions were potential public health nightmares, they were not what bothered him most.

What bothered him was the packing shed, where the freshly picked parsley, radiating heat after being packed in the field in wax boxes, was chilled and sent out. First, workers loaded the boxes on a conveyor belt. The belt traveled through a hydrocooler, a kind of Jacuzzi for vegetables, the job of which was not to clean the parsley but to spray on chilled, recirculated water to keep the produce crisp and fresh-looking and prolong its shelf life. The water came from the municipal system of a nearby village. After the parsley was chilled with the water spray, workers shoveled ice on top from a portable ice-making machine before closing the box, then dumped more ice on each pallet before loading the trucks. The ice was also made from the town water supply.

Later, the CDC team would learn that the farm’s migrant workers, though poor and uneducated, and housed in long narrow barracks surrounded by barbed wire, nevertheless demanded and received bottled drinking water. Soon it became clear why. Vandals frequently disabled the chlorinator for the town water supply, located in a village higher up the mountain. Several times that summer, the chlorinator had been out of service. Usually authorities were able to fix the problem quickly, but a month or so before the investigators arrived, shotgun blasts had damaged the facility beyond repair, and it had stayed that way. The hydrocooler thus sprayed untreated town water onto the produce and recirculated what was left. And though workers say they added chlorine, it probably didn’t help—the dirt from just one box of

parsley could have absorbed all the chlorine that otherwise would have disinfected an entire lot. And that’s probably what happened. The water in the hydrocooler was contaminated either in town or by the newly picked parsley itself. Either way, effluvia from one tainted box could have dripped down on scores of other boxes, spreading Shigella across the lot. Once the boxes were trucked away, Shigella could—and apparently did—spread thousands of miles.

Minnesota officials following the paper trail found that their concurrent epidemic of ETEC traveler’s diarrhea came from parsley on the same farm. And that September, in northern California, more than 300 people suffered severe diarrhea and abdominal pain after eating at a Mexican restaurant. DNA fingerprinting turned up the very same Shigella sonnei from the very same farm—but this time, in the cilantro used in a salsa dip. The hydrocooler was bad news for every green thing that passed through it.

Unlike most outbreaks, all the pieces fit neatly, and yet there was one nagging question: Why didn’t whole parsley cause outbreaks of shigellosis? At the University of Georgia’s Center for Food Safety and Quality Enhancement, scientists tried to find out. They bought a few bunches of raw parsley from a local supermarket, inoculated the leaves with Shigella sonnei, and subjected the herb to a range of preparation methods and temperatures to observe how the bacterium behaved under different conditions. It turned out that mincing the parsley created the perfect medium for the bacteria to grow, because the fluid released from broken tissue cells was loaded with nutrients. While Shigella organisms diminished in the refrigerator, they multiplied with abandon when chopped parsley remained at room temperature. And that’s just what happened at most of the outbreak restaurants: kitchen workers chopped up a large batch of parsley in the morning and let it sit out in a big bowl during the day, grabbing a handful to garnish each dish. With the best of culinary intentions, they had cultivated little Shigella farms.

The parsley outbreak was only one of many warning shots. Increasingly, we are exposed to fruits and vegetables potentially grown in polluted waters or in countries where food safety regulations and work-

ing conditions are less stringent than our own. This was the lesson in 1997, when nearly 300 schoolchildren and teachers in Michigan and Maine were infected with hepatitis A—an often chronic debilitating virus that attacks the liver—after eating frozen strawberries at school. All the fruit had come from a Mexican farm on which strawberry pickers had only a few latrines. The only places they could wash their hands were on trucks that circulated through the fields. With bare hands, the pickers severed the strawberry stems with their fingernails. As the CDC’s Rob Tauxe told Nicols Fox, “Those are the hands that feed you. And it might actually matter whether they are washed or not or whether they have a latrine. If we are interested in the safety of our food, then we have to be interested in the living conditions of the people who handle it.”

We live in a world economy, and there’s no going back. Yet imported produce brings us exotic microbes to which we lack natural immunity. Currently, FDA inspectors examine less than 2 percent of food coming into the United States. And though the agency can seize imports at the border, it can’t bar imported food just because it comes from a country with questionable food safety rules. In fact, in the wake of recent trade agreements, there’s political pressure not to bar such imports, since the action could be interpreted as U.S. protectionism— “nontariff barriers” to free trade—masquerading as food safety policy. Food safety advocates have called for uniform microbial standards for all produce, imported and domestic. It’s a recognition that, as Americans eat more and more from a global plate, the question of whether we have “the world’s safest food supply” is becoming meaningless. We have the world’s food supply.

Just as DNA fingerprinting has supplied crucial evidence to seal criminal convictions, it has revealed surprising connections in a teeming and, until quite recently, unsuspected microbial underworld. Without this technology, the international outbreak of Shigella in parsley—

a perishable produce that quickly passes through the food distribution system—would never have been found. If it hadn’t been so meticulously traced, the hydrocooler in Mexico may have continued to contaminate fresh fruits and vegetables. Public health officials would have been in the dark about this unusual vehicle for transmitting disease, and might not think to ask about it in future outbreaks. More broadly, insights into the fundamental biology of foodborne infection would have been lost. Outbreaks, after all, are public health goldmines. Without recognized clusters of cases and a warm trail of clues, investigators usually can’t find the source of a foodborne outbreak—or ways to prevent the next.

Each day, scores of DNA fingerprints from all over the country fly over PulseNet to a network of public health laboratories dedicated to uncovering foodborne outbreaks. Using a method called pulsed-field gel electrophoresis (PFGE), the labs transform bacterial isolates into a vertical arrangement of bands, from charcoal black to faint gray, representing known sequences of DNA—a kind of genetic bar code. The technique is well suited to investigations of foodborne bacteria because, even after many generations spreading and dividing in both animal and human hosts, bacteria descended from the same parent still contain virtually the same genetic material. Lookalike PFGE patterns in two different samples suggest that the bacteria came from a common source, such as a widely distributed contaminated food. The fingerprints are entered into an electronic database at state or local health departments, and are also transmitted to the CDC, where they enter a closely monitored computer. Currently, PulseNet labs can fingerprint E. coli O157:H7 as well as other forms of E. coli, Shigella sonnei, nontyphoidal Salmonella, Listeria monocytogenes, Campylobacter jejuni, and others. But in the next few years, gene chip technologies will truly open the frontier, enabling labs to diagnose dozens of bacterial, viral, and parasitic pathogens in a single test.

Because it reveals camouflaged epidemics triggered by widely distributed products with spotty and low-level contamination, PulseNet is perfectly suited to the foodborne outbreaks of the future. Until recently, the scattered cases of illness caused by these subtle outbreaks

were deemed “sporadic”: separate and random, with no known link. PulseNet connects the dots. And because of PulseNet’s sensitivity, investigators are finding smaller and smaller clusters of cases, and more of them—so many, in fact, that CDC epidemiologists are, like hurricane-watchers, giving the outbreaks names. This doesn’t necessarily mean disease is increasing, but that the ability to detect it is improving. Between 1995 and 1999, for instance, the median size of a detectable E. coli O157:H7 outbreak dropped from 27 cases to 9. From a public health standpoint, that’s good, since the more clusters officials can pick up, the more chances they have to follow leads back to the cause.

PulseNet’s leads have also startled public health sleuths. What before may have looked like a single outbreak sometimes proves to be composed of several separate and concurrent outbreaks. And what used to seem like unlinked sporadic cases now sometimes turn out to be part of a dispersed epidemic with a single cause. “We had this old belief that if we were good public health people, we could get in there when the fire started and we could save the town,” says Mike Osterholm. But with the old technologies, health officials noticed only the raging fires wiping out city blocks. Today, with DNA fingerprinting, they can see all the little house fires—too many fires, in fact, to put out. State and local health departments, traditionally short-staffed and underfunded, can’t possibly put out or even investigate every conflagration.

PulseNet has also spurred an almost philosophical shift in public health. In a postmodern turn of phrase, the CDC’s Eric Mintz refers to a new landscape of “complex meta-outbreaks,” in which many more foodborne illnesses are tied to sources of contamination far back in time and space. DNA fingerprinting may reveal a dense web of microbial connections between farm animals and produce operations and food processors and the tens of millions of Americans each year who get sick from what they ate. After all, foodborne illness doesn’t happen by accident. “I used to hold that there was no such thing as a sporadic case of an infectious disease; there were only patients who were not part of a recognized outbreak,” Mintz says. Now, he’d extend this line of thinking even further. “Are there really any ‘sporadic’ outbreaks?

Or are all outbreaks related to other meta-outbreaks that we are now more able to discern?”

Even if investigators were able to tie together individual cases and outbreaks in a grand theory of everything, they wouldn’t necessarily be able to act on that knowledge. Few investigations end as neatly or as gratifyingly as the case of the Shigella-ridden parsley. Sometimes DNA fingerprinting raises more questions than it answers, or presents bureaucracies with challenges that they either won’t or can’t face. Unlike in criminal investigations, DNA fingerprints aren’t enough to close a case of foodborne illness, since there is a tiny chance that identical bacterial fingerprints may come from unrelated sources. Epidemiologists must also prove that the patients all ate a particular food and, for liability reasons, federal regulators must prove that the product was indeed contaminated.

The sometimes tricky business of closing a PulseNet case was underscored during an outbreak in the United States in late 1998. That November, the CDC received calls from several states reporting an increase in Listeria monocytogenes cases. Named for the British surgeon Joseph Lister, the father of modern hospital antisepsis, Listeria— a small bacterium that tends to line up side by side, like a regiment of recruits—is almost impossible to eradicate once it makes its home in a meat processing plant. Ubiquitous in the environment, it is found in soil, water, sewage, decaying vegetation, and many species of birds and mammals. Reports of Listeria are usually the kickoff for frustrating and ultimately fruitless investigations because the organism has a long incubation period; the lag time between eating a contaminated food and developing symptoms can be one to two months. That makes gathering food histories difficult, since most people can’t remember what they ate yesterday, let alone eight weeks ago. More perplexing, the symptoms of listeriosis don’t always suggest foodborne infection. In fact, not until 1981—after an outbreak in Canada killed 41 percent of those infected, and 34 women suffered stillbirths or miscarriages or gave birth to ill infants—did researchers even realize Listeria was transmitted by food. The victims had eaten commercially prepared cole slaw made from cabbage grown on a farm where listeriosis had killed

sheep. If a mother’s infection spreads to the fetus through the placenta, Listeria can cause meningitis in infants and newborns. Listeria infection in the nervous system causes a severe headache, stiff neck, loss of balance, and convulsions. All told, Listeria monocytogenes kills 1 in 5 people it infects, making it one of the most deadly foodborne agents known. Each year in the United States, more than 2,500 persons fall ill and more than 500 die. But while between 2 and 7 percent of processed meats carry Listeria, some strains may be less virulent than others, which is why there aren’t constant Listeria outbreaks. “Chances are most Americans eat Listeria monocytogenes at least once a week,” says the CDC’s Paul Mead, “and yet they don’t get sick that often.”

The 1998 case was one of the deadly strains. By late November 1998, the death toll was mounting. Investigators had no clue where the disease was coming from. Questionnaires eventually ferreted out what the victims had in common: hot dogs. But which brand? Many had been mentioned. Here, a lucky break—also the legacy of DNA technology—pointed investigators toward the answer. That fall, Cornell University happened to have taken genetic fingerprints from pieces of smoked chicken that had sent several members of a New York family to the hospital. The DNA fingerprint perfectly matched the hot dog outbreak strain—and both, it turned out, came from Bil Mar Foods, a division of the Sara Lee Corporation, the biggest purveyor of packaged meats in the country.

By mid-December, 15 people scattered across the country had died in the outbreak and scores were ill—though with the slow trickle of reports from state health departments, the CDC wouldn’t be able to add up the numbers for weeks. Ultimately, the agency calculated that the outbreak killed at least 21—15 adults and 6 stillbirths or miscarriages—and made more than 100 sick in 22 states. But the CDC and USDA tussled over whether there was enough evidence to implicate the company—while CDC insisted there was, USDA officials fretted about the legal consequences of wrongly accusing Sara Lee. Understandably confused by the government’s mixed signals, Sara Lee officials called Mike Osterholm, then the Minnesota state epidemiologist,

who had been at the helm of many high-profile foodborne illness investigations. Osterholm was convinced the company should carry out a recall. As he explained to the Washington Post, in a typical metaphorical flourish, “I said to Sara Lee that this was like driving down a highway and seeing a herd of deer. You can put on the brakes now and you’ll still hit them. Or you can hit them at full speed and the deer will come through your windshield.”

On December 22, Sara Lee recalled 15 million pounds of hot dogs and deli meats. But the USDA, contrary to its own policies, did not deliver a timely warning to the public; instead, it allowed Bil Mar to issue a weakly worded press release that did not mention the serious nature of the illness, or the deaths. A few days later, the CDC team isolated the Listeria strain found in a package of unopened hot dogs in the refrigerator of one of the Michigan patients—a woman who had given birth to a child with sepsis. The strain matched the bacterium that was quietly killing Americans.

But the government’s earlier timidity had deadly consequences. In the midst of the holiday season and as presidential impeachment dominated the news, Americans never really took note of the outbreak. In Columbus, Ohio, Lisa Lee and her fiancé, unaware of the recall, continued eating Sara Lee deli meats. Lee was in the fifth month of her first pregnancy, expecting twins. That January, feeling feverish and wretched, she went to an emergency room, where a doctor diagnosed her with flu and sent her home. A few days later, she returned with a high fever. After unsuccessfully trying to stop her contractions and prevent miscarriage, doctors induced a long labor. Lee gave birth to twins, a boy and girl, both stillborn. “If we knew it as soon as Sara Lee knew it,” she told an interviewer, “we might have had a chance.”

As it happened, employees at the Bil Mar plant did know about the contamination, long before the recall. The plant had had trouble with condensation in the hot dog production room. When franks are cooked to a high temperature and then chilled, steam rises to cooling elements in the ceiling. Those cold, moist conditions make a perfect home for Listeria—a psychrophilic, or cold-loving, bug—which thrives on walls and ceilings and can survive for years on biofilms and metal

surfaces. When the steam condenses, it rains down on whatever happens to be sitting on the production line, washing Listeria along with it. On the July 4th weekend, Bil Mar workers had ripped out cooling units from the ceiling and, because the units were too big to push out the door, cut them up with a chain saw before forklifting them away. Immediately afterwards, routine bacterial samples came back positive. The construction probably raised dustborne Listeria that drifted through the factory to the deli meat section, under the same roof but three acres away. From the plant, Listeria was dispatched across the country. In response to this dramatic foodborne epidemic—the most lethal in the United States in 15 years—the USDA in 2000 took steps to require Listeria testing in processing plants. In 2001, Sara Lee pleaded guilty to a misdemeanor charge that it had produced and distributed tainted meat. It agreed to pay $4.4 million to settle civil and criminal charges.

The very things that make America’s food system extraordinary— its mammoth production levels, its stunning variety of imports, and its low cost—are what compromise safety. Mass production spreads risk, imports bring global pathogens to our doorstep, and historically low cost makes farmers and regulators reluctant to pursue safety measures that could raise prices. It’s not the ideal way to cultivate the world’s safest food supply—a claim that food safety advocates say is overblown anyway. Denmark, Sweden, and the Netherlands, for instance, have taken far more vigorous steps to eliminate pathogens.

Even if America’s were the safest, it’s not safe enough. So whose fault is that? Throw together all the parties involved in food safety— scientists, company officials, regulators, advocacy groups, victims and their families—and the discussion inevitably boils down to this: How much responsibility for preventing foodborne illness belongs to the individual and how much to government and industry? Those who lean toward individual responsibility point out that it’s often mistakes close to home—either yours or someone else’s—that inflict the

damage. Most cases of Campylobacter, for instance, simply come from undercooking poultry or letting the raw bird or its juice touch another food. Most cases of E. coli O157:H7 come from eating ground beef heated only to pink or pallid gray. “Raw meats are not idiot-proof,” says USDA microbiologist Nelson Cox. “They can be mishandled and when they are, it’s like handling a hand grenade. If you pull the pin, somebody’s going to get hurt.” But while some might question the wisdom of selling hand grenades in supermarkets, Cox believes that blame grows less actionable the farther back you go in the food chain. “Who are you going to sue?” he asks. “Are you going to sue Kroger? Or are you going to sue the man who transported it? Are you going to sue the man who processed it? Or the one who grew it? Or had the breeder flocks? I think the consumer has the most responsibility but refuses to accept it.”

Admittedly, food abuse is rampant. Surveys show that more than one-quarter of respondents don’t wash cutting boards after cutting raw meat, nearly a quarter prefer their hamburgers pink, and half eat raw or undercooked eggs. The faster we forget the lessons of an agrarian culture, and the more we rely on packaged and microwavable food, the more we are apt to mishandle raw ingredients in our own kitchens. Moreover, changes in work and family life have led us to cook less and eat out more—Americans spend half of their food dollar away from home, and nearly 14 percent on fast food—exposing us to the ministrations of young and often even more naive food handlers. A 1999 study from Los Angeles found that large restaurants—those with more than 400 seats—were more than seven times as likely to have a complaint lodged against them as a restaurant with fewer than ten seats. It

wasn’t just because the restaurants had more customers, but because they had varied menus, creating more chances for cross-contamination behind those swinging kitchen doors.

Inexperienced kitchen help who may not receive training in sanitation or hygiene are also perfectly positioned to broadcast their own foodborne ailments. Because wages in these jobs are low, and insurance and sick leave nonexistent, “people are encouraged to work while they’re sick,” says Kirk Smith, who supervises foodborne disease investigations at the Minnesota Department of Health. In addition, many food workers emigrate from countries where intestinal infections are endemic. For all these reasons, employees suffer unusually high levels of enteric disease. While up to 0.5 percent of individuals in the general population may carry an intestinal parasite, in restaurants that have triggered foodborne outbreaks, up to 18 percent of food handlers have been shown to suffer intestinal infections. The public health literature is replete with reports of kitchen workers sowing foodborne epidemics, from Shigella to Campylobacter to hepatitis A viruses. Though shellfish from polluted waters are a major source of Norwalk virus, a nasty ailment that begins with sudden projectile vomiting, so are infected kitchen workers who are less than meticulous—such as the bakery worker in Minnesota in 1982 who used his bare hands and arms to stir 76 liters of butter-cream frosting, an unconventional culinary technique that made 3,000 pastry lovers sick. In 1988, the largest home-grown epidemic of Shigella came to light when 21 players on the Minnesota Vikings football team suddenly became ill after a charter flight to Miami, a trip on which they had partaken of cold meat sandwiches. After media coverage of the outbreak, hundreds of calls poured in to the Minnesota Department of Health from other air travelers complaining that they, too, had gotten sick. Investigators ultimately tracked the outbreak to 219 airline flights depositing travelers in 24 states, the District of Columbia, and four other countries—jaunts on which as many as 35,000 travelers ate contaminated sandwiches. One or more employees in the flight kitchen—who couldn’t afford to lose a day’s pay—had worked while sick with diarrhea from shigellosis.

It’s these hidden, unsuspected sources of food poisoning—from

long-contaminated meat and produce, from infected food handlers, from unpasteurized apple and orange juices—that have persuaded food safety advocates that fallible humans simply can’t act as the bulwark against dangerous pathogens. In the United States, at least 12 federal agencies have a hand in food safety, fiefdoms that sometimes work in efficient synchrony but often in a kind of all-thumbs opposition. Though the system has become a sprawling mess, it did make sense during the era of malfeasance when it was created. In 1906, Congress passed the Pure Food and Drugs Act, which established the duty of the federal government to regulate foods other than meat and poultry, and to prohibit the interstate sale of food that was misbranded or adulterated with chemical preservatives—a structure that’s now the Food and Drug Administration. And in response to Upton Sinclair’s 1905 novel The Jungle, which graphically depicted Chicago’s meatpacking industry, lawmakers passed the Meat Inspection Act of 1906. That law set sanitary standards for slaughter of animals and for meat sold in interstate commerce and led to daily inspection in slaughterhouses, using “organoleptic” means—sight, smell, touch—to ferret out problems.

Back when the nineteenth century became the twentieth, of course, “clean” didn’t mean free of microscopic pathogens, since many were not known at the time. Because of this technological lag, the public and the courts later came to consider disease-causing organisms in meat and poultry an unavoidable risk. As late as 1974, in a case in which the American Public Health Association sued the U.S. secretary of agriculture for violations of the Wholesome Meat Act, a U.S. court of appeals endorsed this fatalism, writing that Salmonella and other organisms were “inherent” in meat and that it was the consumer’s responsibility to work around those hazards: “American housewives and cooks normally are not ignorant or stupid and their methods of preparing and cooking of food do not ordinarily result in salmonellosis.” Not until 1994 did a district court break a new precedent, declaring that “E. coli is a substance that renders ‘injurious to health’ what many Americans believe to be properly cooked ground beef” and therefore “fits the definition of an adulterant.”

Shaped by case law and political wire-pulling, the USDA and FDA have dramatically evolved from their origins. The USDA is now responsible for monitoring meat, poultry, and commercially processed egg products, with a policing force known as the Food Safety and Inspection Service (FSIS). The FDA is charged with ensuring that all other foods are safe, nutritious, sanitary, wholesome, and honestly labeled. This jigsaw puzzle sometimes doesn’t fit. For example, while the FDA oversees shell eggs, the USDA has authority over egg products. While the FDA oversees plants producing cheese pizza but rarely inspects them, the USDA has jurisdiction over plants producing pepperoni pizza and inspects them every day. And while the USDA is responsible for meat and poultry safety, its inspectors can look only for microbes that cause animal—not human—disease. The two agencies also share monitoring of imported foods, again with very different approaches. While the USDA must certify an exporting country’s meat and poultry inspection program, the FDA has no such power and instead relies on a tiny staff to do physical inspection and chemical analysis of food at ports of entry—inspections that can’t possibly keep pace with the huge rise in imports. The FDA inspects less than 1 percent of the foods and ingredients imported from other countries.

Today, the centerpiece of the federal food safety system is the HACCP program—for Hazard Analysis Critical Control Point. HACCP (pronounced has´-sip) is a system of process control that evolved out of methods designed to protect astronauts from food contamination. The model is an eminently rational one. It identifies critical contamination points all along the path from farm to fork—in shipping, processing, wholesaling, retailing, and kitchen handling— eliminates those hazards, monitors the effects, and then sets even more stringent standards. But it may hand over too much responsibility to industry to police itself. In 2000, the General Accounting Office, a research arm of Congress, issued a report recommending that the government tighten enforcement of sanitation standards in meat and poultry plants, and give the USDA the authority to levy fines against plants that don’t meet these standards. The GAO also estimated that 85 percent of foodborne infections comes from fruits, vegetables, seafood,

and cheeses: items regulated by the thinly staffed FDA, an agency with only a tenth of the inspection force of the USDA.

Critics of American food safety policies have insistently called for a single independent government agency to replace the current patchwork of bureaucracies and fiefdoms that are often at cross-purposes. CDC officials privately grumble about the USDA’s and FDA’s foot-dragging and the bureaucratic hurdles that spring up the instant an outbreak investigation is launched. “Until a few years ago, the attitude of the Department of Agriculture was that if a pathogen didn’t make the animal sick, it wasn’t their business—it was your business,” says Carol Tucker Foreman, a former USDA official who now heads the Safe Food Coalition, an advocacy group in Washington, D.C. “You have diffuse authority and a political structure that says: Do not impose any cost on the farmer. You don’t go tell the American farmer that he has to rebuild his stock pens, redo his water supply, alter completely the way he houses and ships animals to market. Look at the people who make up the House and Senate agriculture committees: Nobody’s on those committees because his or her first interest is food safety. They’re on those committees because their first interest is to make sure farmers make a profit.”

Meanwhile, industry and regulatory officials counter that CDC investigators operate with a naive, narrow-minded zeal. Egg industry officials, for instance, complain that the Clinton administration and the CDC focused attention on eliminating Salmonella enteritidis in eggs only because SE outbreaks almost always originate with that single food, making it a conveniently neat target—unlike E. coli O157:H7 or other forms of Salmonella, which pop up all over. “Some of the people that we interact with from CDC . . . say things . . . without ever having personally been on a farm,” observes the USDA’s program leader for the National Animal Health Monitoring System, Nora Wineland. “I would take it with a very large grain of salt.” So deep is the antagonism at times between the federal agencies involved in food safety that Wineland has even wondered aloud whether the many cases of disease that CDC officials are convinced are foodborne actually come from food at all. It’s perfectly plausible, Wineland says, that someone could

get a Salmonella infection by preparing food on a kitchen counter where the family cat may have just been prowling. “And where’s that cat been?” Wineland asks. “When you try to think about where people could pick up these bugs, the possibilities in my mind seem nearly endless.” It’s an argument that makes CDC investigators see red.

The finger-pointing goes in all directions. Public health officials blame farmers and industry for sanitation problems that they believe perpetuate the chain of infection. Farmers blame slaughterhouses and processing plants for contaminating their products. Microbiologists studying animal diseases insist that fresh produce is the biggest threat today. And the tension between blaming consumers for foodborne disease and blaming corporations and government never goes away. “It’s the blame game,” says Mike Osterholm, adding that rhetoric doesn’t go far. “Today, if I gave some of these advocacy groups a billion zillion dollars and told them to raise cattle and bring them to market and get them to the consumer without E. coli O157:H7, they couldn’t do it. And if I gave a billion zillion dollars to a meat packing plant and told them, ‘You can’t have any Listeria, period—zip, zero,’ they could do a lot to minimize it but they couldn’t reduce it to zero.” Osterholm wonders, in fact, if the problem could ever be dispelled at the source. “The bottom line is, if everybody did their job perfectly well, given the level of technology that’s readily available, you’d still have problems.” Food recalls, he adds, almost always come too late. “It’s like shutting the barn door after the cow is out.” Unlike many public health experts, Osterholm has pushed hard for a last-stage technical fix: irradiation of ground beef, poultry, and eventually of fresh produce, all of which he likens to the commonsensical pasteurization of milk.

Ultimately, preventing foodborne infections will take scientific and political courage—sparked, most likely, by personal stories. In 1993, Alexander Thomas Donley, age six, was one of four children who died from eating a Jack in the Box hamburger. E. coli O157:H7 first made him curl into a fetal position from abdominal cramps. Then, one after another, his organs failed. Screams of pain were followed by silence as toxins liquefied his brain. He suffered tremors and delusions and finally a massive seizure. His body swelled as his kidneys shut down. “I

was so horrified and so shocked and so angered by what happened to him,” says his mother, Nancy Donley, now the president of Safe Tables Our Priority, or S.T.O.P., a Chicago-based advocacy group. “I had no idea that there was any problem in our food supply. I loved my child more than anything in this world. And then to find out that he died because there were contaminated cattle feces in a hamburger. And to find out that had been recognized as a problem for a while. Why hadn’t it been fixed?”

Donley believes in the value of teaching consumers to handle food safely, but she doesn’t think education is enough. The best way to prevent foodborne illness, she says, is to mount a strong attack on the farm and then vigilantly follow up every step along the path to the table. It’s an approach that goes against the grain of many political interests. Not long ago, testifying before a congressional subcommittee, a senator reminded her, as have countless industry and government officials over the years, that the United States has the safest food supply in the world. Donley stared at him and said, “Senator, I beg to differ with you.”