Stealth Attack

I. Flight of the Bees

An unprecedented phenomenon is plaguing American agriculture: Honeybees are disappearing. Starting in 2006, the population of honeybee hives around the United States has dropped by roughly 30 percent each year, an event now known as bee colony collapse disorder. Some individual colonies have lost up to 90 percent of their bees.

Bees are essential to farming. The United States Department of Agriculture (USDA) estimates that bee pollination contributes more than $15 billion annually to crop values.

The plight of the bees may also be incredibly important to human health. In fact, says Chensheng Lu of the Harvard School of Public Health, bees could be serving as a canary in the coal mine.

In the summer of 2010, Lu, an associate professor of environmental exposure biology, began a 14-week field experiment. At each of four different sites, he set up five hives and supplied the food most commonly (and cheaply) used to sustain bees over the winter months—high-fructose corn syrup. In four of those five hives, the corn syrup was spiked with varying amounts of the pesticide imidacloprid. The fifth hive served as a control; nothing was added to the corn syrup. The 20 hives buzzed with about 80,000 bees each.

Imidacloprid is currently one of the most widely used pesticides in the world. A neurotoxin, it causes swift death by paralysis. It blocks specific receptors in an insect's central nervous system so that nerve cells cannot take up the neurotransmitter acetylcholine, which is essential for muscle function.

Since 2004 the chemical has been sprayed on the vast majority of corn crops, and especially genetically modified corn seed, to eliminate such insects as aphids, termites, and beetles. Corn is by many multiples America's largest farm product. Studies have shown that imidacloprid makes its way from the germinating seed into the growing plant and then into the corn kernels.

Lu suspected that the timing of the disappearance of America's bees was significant. Beekeeping is an ancient practice; little has changed over time except the use of high-fructose corn syrup as winter food in place of more valuable honey. It is made from sugars extracted from corn kernels.

Thirteen weeks after imidacloprid was first added to their food supply, Lu's bees began dying in such significant numbers that two hives were completely dead. Within 23 weeks, 15 of the 16 hives exposed to the insecticide were empty, and the hives getting the highest dosages died off the fastest. But bees also died on a dosage of 20 parts per billion, which is less than they would encounter when foraging crops sprayed as seed. Only one control hive died—from an outbreak of dysentery.

"The results from this study show a profound and devastating effect of low levels of imidacloprid in high-fructose corn syrup on honeybee colonies," Lu and his colleagues reported earlier this year in the Bulletin of Insectology.

Virtually simultaneously, other studies reported equally disturbing effects. One found that the pesticide disrupts performance of the so-called waggle dance, a midair maneuver bees engage in to make them visible to other hive members and alert them to the location of good food sources. Worker bees wind up bringing back far less food to the colony. And as a result, colonies produce far fewer queen bees. Imidacloprid exposure also disrupts the internal navigation system of bees. They can't find their way home after foraging.

If Lu's study of imidacloprid in bee colony collapse disorder proves anything, it's that acute pesticide toxicity is scarcely the only threat to health. Exposure to high enough doses of imidacloprid indeed kills through acute toxicity. Badly exposed bees don't even make it back to the hive. But what no one predicted was that extremely low doses would also be deadly. Borne by stealth in high-fructose corn syrup, they get directly into the hives and create the bigger threat of mass extinction, far downstream from contact with treated seeds.

The levels of exposure to imidacloprid were so low in Lu's study that he draws a sobering conclusion: There's really no safe dosage of imidacloprid for bees. That's an opinion echoed by the governments of Germany, France, and Italy, which have all suspended use of the insecticide. The EPA has jurisdiction over pesticide levels in foods, the USDA monitors residue levels in meat and eggs, and the Food and Drug Administration enforces standards for produce and the rest of the food supply. But no one, Lu notes, monitors residue levels of imidacloprid in high-fructose corn syrup—because it is not marketed directly to consumers.

"This study points out that sublethal, low dosages of pesticides are causing adverse health outcomes," the Harvard biologist explains. What's more, he says, the results indicate that, just like the bees in his study, people may be unwittingly exposed to small amounts of imidacloprid that are subtly undermining human health. The agrochemical industry and regulators do not accept evidence from insects as proof that human health is at risk. "Without an abundance of human data," says Lu, "this pesticide can still be used for another 20 or 30 years."

II. Cognitive Concerns

There is very little data on human exposure to imidacloprid. Nevertheless, a spate of recent epidemiological studies establishes troubling connections between low-dose exposure to a variety of common pesticides and conditions ranging from attention deficit hyperactivity disorder in children to neurodegenerative conditions such as Parkinson's disease in later life.

The modern synthetic pesticide era got underway in 1945, when DDT was released for agricultural use and crop yields skyrocketed. By 2007, over 5 billion pounds of synthetic pesticides were sprayed, dusted, or coated on crops worldwide, according to the EPA. The U.S. uses 22 percent of the total to produce 4.3 percent of the world's agricultural output.

Aside from direct contact in homes and gardens, the most common route of pesticide exposure is via the food supply, through residues on, and in, fresh or processed fruits and vegetables, on corn and wheat products, and in drinking water. People living in agricultural areas face additional exposure by working directly with pesticides and inhaling drift from fields where they are sprayed. Each year, physicians diagnose 10,000 to 20,000 cases of poisoning among the 2 million farm workers in the U.S.

The EPA sets human "tolerances" for pesticide exposure through tests of animal models. Its working maxim is "the dose makes the poison." Laboratory rats are exposed to high doses of a chemical and examined for physical or neurological evidence of toxicity. The dosage is then lowered to the point at which no effects are discernible. That dosage, divided by a safety factor, yields the acceptable exposure level for humans. The USDA's most recent annual survey, for 2010, found that only 0.25 percent of 12,845 samples of drinking water and food contained pesticide residues above tolerance limits set by the EPA. The report concluded that pesticide residues pose no safety risk through food.

"In current toxicological testing for pesticides, in order to meet regulatory standards, we focus more on high levels than on lower levels," Lu explains. "But low levels are more relevant to human exposure on a daily basis." And as with bees, effects may be occurring at "nonobservable" levels.

With DDT banned by 1972 because of its persistence in the body and the environment and damage to an array of wildlife, the agricultural industry turned to a class of chemicals known as organophosphates. Far less persistent than DDT, hanging around the body a mere three to six days, they are, however, far more instantly toxic (think sarin). Initially developed as nerve agents during World War II, they do their gruesome work by blocking acetylcholine neurotransmission. By 1990, organophosphates totaled 70 percent of all agricultural insecticides; they are commonly used on cotton, alfalfa, and oranges.

Their ability to attack the human nervous system has made the organophosphates obvious candidates for research on the health effects of pesticide exposure. Animal studies have documented a portfolio of effects. Mice and rats exposed prenatally exhibit not only lower birth weights and motor problems but reduced brain volume in critical areas (notably the cerebellum and the hypothalamus) and cognitive difficulties (seen as increased time to navigate a maze). In 1993, the National Academy of Sciences declared that "exposure to neurotoxic compounds at levels believed to be safe for adults could result in permanent loss of brain functions if it occurred during the prenatal and early childhood period of brain development."

Mounting concern about children's exposure to organophosphates led the EPA to phase out residential use of the pesticide chlorpyrifos beginning in 2000, and manufacturer Dow Chemical Company withdrew it from the residential market in 2002. Chlorpyrifos is still widely used agriculturally, notably on corn, grapes, oranges, apples, and almonds.

A 2010 study published in the journal Pediatrics by researchers from Harvard and the University of Montreal found that current exposure to a group of organophosphate pesticides increases the risk a child will develop attention deficit hyperactivity disorder. The researchers piggybacked their study onto data collected from 2000 to 2004 through the National Health and Nutrition Examination Survey, a periodic survey of 5,000 people designed to check the health status of U.S. children and adults. The survey collects spot urine samples, which the Pediatrics team checked for levels of organophosphates and their breakdown products.

"Almost everyone—96 percent of kids—had detectable levels of organophosphates in their urine," reports Montreal researcher Maryse Bouchard. Children 8 to 15 years old whose urine contained above average concentrations of organophosphate metabolites had double the odds of meeting the diagnostic criteria for ADHD. "Since organophosphates are supposed to be cleared from the body in a few days, and yet we measured them in everybody, it means that we are exposed repeatedly. Everybody had a little dose everyday, probably," presumably through pesticide residues on foods.

Three studies of organophosphate exposure recently reported in the journal Environmental Health Perspectives found measurable cognitive deficits in the offspring of women who were pregnant between 1999 and 2000 in New York City and California, although all exposure levels were in the acceptable range set by the EPA. Women on both coasts had used organophosphate pesticides in their homes; further, the Californians are farm workers from Salinas Valley. The children have undergone cognitive testing throughout childhood. Those whose mothers registered higher levels of exposure during pregnancy—in tests of cord blood or urine—have demonstrated significant deficits in perception and memory compared with those whose mothers tested relatively pesticide-free.

Among 329 children from farm-worker families, most recently tested at age 7, researchers found a seven-point difference in IQ between those whose mothers recorded the greatest exposure and those whose mothers had the lowest. Montreal's Bouchard, a member of the research team, offers a clear perspective on the deficit: Pregnant women are cautioned to avoid eating fish like tuna because it may contain traces of mercury. Prenatal exposure to mercury is linked to a loss of one IQ point in children. "In the world of environmental health," she says, "seven points is a big deal."

Researchers at Mt. Sinai medical center have been testing 404 mother-child pairs in New York. Beginning at 1 year of age, children of high-exposure mothers showed difficulty with spatial reasoning and visual processing, as measured in exercises with blocks and pictures. The cognitive deficits persist at age 9, especially in those children whose mothers have a common gene variation that compromises their ability to break down organophosphates. Thirty percent of the women have that gene variant.

A third study documented a three-point drop in IQ, marked by deficits in working memory among the most exposed of 265 children in New York City. All the exposure levels were within the EPA's acceptable range. The findings, observes Virginia Rauh of Columbia's Mailman School of Public Health, "have implications for learning, and they're not good."

In mid-July, the EPA, concerned about the IQ effects of chlorpyrifos on children in agricultural communities, issued new limits on the amount of the chemical that can be applied to fields and orchards—but stopped short of an outright ban.

III. Hormone Disruption

The organophosphates under study are not exerting their effects on cognition through the established pathway of harm, disruption of acetylcholine neurotransmission. Rather, some other mechanism is at work.

Columbia's Virginia Rauh sees a tantalizing clue in her most recent findings, reported in Proceedings of the National Academy of Sciences. They show that the pesticide alters many brain structures in a strange variety of ways. She and her team carried out MRI scans on 40 of the New York City children, at ages 6 to 11, who had been exposed to the organophosphate chlorpyrifos in utero, 20 at low levels, 20 at higher levels.

Overall brain volume was unaffected. But the cerebral cortex, the site of information processing and working memory, showed generalized thinning, although some areas were enlarged. The thinning of the prefrontal cortex may explain the IQ drop found among the high-exposure children.

Significantly, there were signs of sexual reversal in the size of some cortical structures. Areas that are normally larger in girls were larger in highly exposed boys, notably areas related to self-awareness and sensory information processing. "We don't know how these brain anomalies came about," says Rauh.

It may be that at low levels, chlorpyrifos, in addition to being a neurotoxin, disrupts the normal functioning of endocrine hormones, which are critical for early organ development as well as later for sexual and other functioning. Chlorpyrifos and its organophosphate siblings in many ways mimic the hormones estrogen and testosterone.

Laura Vandenberg of Tufts University's Center for Regenerative and Developmental Biology reviewed more than 800 studies of the effects of such agents. "Chemicals that mimic hormones, even in low doses, can have profound effects on the biology of our bodies," she reports. And those effects cannot be predicted by the actions of the same agents at higher doses.

Most research on the endocrine-disrupting action of pesticides has been done in animals. The best studied agent is atrazine, an estrogen-disrupting herbicide widely used on corn, sugar cane, and sorghum. Runoff from crops sprayed with atrazine demasculinizes male frogs, turning them into hermaphrodites bearing both male and female reproductive organs. In its most recent annual survey, the USDA found traces of the weed killer in more than 99 percent of nearly 300 samples of drinking water.

Chlorpyrifos is designated a mild endocrine disruptor. Vandenberg finds it has antiandrogenic effects; it interferes with male sex hormones, such as testosterone, responsible for the development of male characteristics from sperm production to large-muscle mass. Mice exposed to low doses of chlorpyrifos during gestation had decreased levels of thyroid hormone five months after birth. Thyroid hormone, Vandenberg observes, is critical to brain development.

In the absence of human evidence, Columbia's Rauh isn't quite ready to assign to chlorpyrifos the sex-related brain anomalies she observed. Still, she admits, "it might be associated with some reversals of normally expected sexual differentiation." Animal evidence suggests the brain changes will not be modified by puberty.

IV. Additive Effects

Virginia Rauh characterizes as "modest" the effects she has found so far of relatively low levels of chlorpyrifos exposure. But she quickly points out that we rarely encounter chemicals like chlorpyrifos in isolation. Although the EPA tests for and establishes tolerable limits for each chemical by itself, "it's very unrealistic to suggest that people encounter single exposures," Rauh says. "Children who have high exposures to one chemical likely have simultaneous exposure to another chemical, or multiple chemicals." The effects of residues of organophosphates commonly found on the same foods are, at the very least, additive.

Further, chemicals that act on the body through differing mechanisms can jointly have an amplified impact on human health. A 2009 study reported in the American Journal of Epidemiology looked at 368 people who had been diagnosed with Parkinson's disease, the neurodegenerative condition characterized by disabling tremors, between 1998 and 2007. All 368 lived within 550 yards of farm fields that had been sprayed with pesticides for at least five years prior to their diagnosis.

Guided by animal studies, the researchers focused on exposure to two agents, the herbicide paraquat and the fungicide maneb. Mice exposed to both pesticides early in life display Parkinson's-like neurodegeneration as they age. The human residents most likely to have been exposed to both pesticides had a 75 percent greater than normal risk of developing Parkinson's disease. Those whose condition was diagnosed before age 60 had double the normal risk of disease from earlier exposure to just one of the chemicals—and four to six times the odds if they had been exposed to both.

By differing paths, both agents affect the brain's dopamine neurons, the cells that die off in Parkinson's patients. Paraquat is related to a compound known to interfere with the mitochondria, or energy center, of such cells. Maneb interrupts the process cells employ to eliminate defective proteins. The accumulation of specific proteins in dopamine-sensitive neurons is a hallmark of brains afflicted with Parkinson's.

Pesticides are a necessity of modern life. They help ensure an adequate food supply. Therein lies another dilemma.

"If you remove one chemical or class of chemicals from the list of pesticides growers can use, they'll replace it with something we know even less about," observes Brenda Eskinazi, the epidemiologist at the University of California at Berkeley who led the Salinas Valley study. She points directly to pyrethroids, which are synthetic versions of naturally occurring pesticides (pyrethrins) made by chrysanthemums. Pyrethroids have been growing in popularity since the 1990s, amid increasing concerns about the human toxicity of organophosphates. Pyrethroids are now components of more than 3,500 insecticide products, especially household ones. They alter nerve cell membranes to create paralysis and death of insects. "We know almost nothing about pyrethroids in humans," Eskenazi notes.

As the scientific community gains a better understanding of the neurotoxic and hormone-disrupting effects of low-dose pesticides, one theme recurs: The current ways of defining exposure are no longer sufficient. "Tying everything to non-observable doses—deciding the safe dose based purely on the level that doesn't cause a visible effect on a mouse or rat—is not going to protect human or even ecological health," says Lu. "The system we have been using since the birth of the EPA needs to be completely overhauled."

Dosage no longer seems to matter. One way or another, poison is still poison.

Combatting Pesticides on Produce

For the vast majority of Americans, the most common source of exposure to pesticides is residue on food, according to the Pesticide Action Network of North America. Long after crops are sprayed, trace amounts of pesticides or their breakdown products persist on the surface—and in the flesh—of many fruits and vegetables.

The simplest way to minimize exposure is to wash fruits and vegetables thoroughly, whether or not you will eat the skin. According to Berkeley's Brenda Eskenazi, even though you don't usually eat the skin of an orange, not scrubbing it could still result in the transfer of pesticide residues to the fruit flesh during peeling or slicing.

Although it is not an option for everyone, buying organic produce, which is grown without the use of traditional pesticides, can significantly curb exposure. A 2006 study by Harvard's Chensheng Lu, then at Emory, showed that detectable levels of organophosphate metabolites in the urine of 23 schoolchildren vanished immediately after organic food was substituted for their regular diet.

At the very least, consumers should aim to buy organic for the "Dirty Dozen" fruits and vegetables that the Environmental Working Group consistently finds carry the greatest pesticide burden. Apples, celery, and bell peppers regularly top the list.

The worst choice a person can make is to avoid eating fruits and vegetables altogether, says Eskenazi, especially during pregnancy. "Those fruits and vegetables are what your child needs to make the brain grow properly."