February 17, 2004
Models help estimate children's exposure to toxins
By Geoff Koch
For almost 10 years, Stanford's Jim Leckie and his students have been successfully collecting immense amounts of data, writing original software and building sophisticated statistical models -- all to begin to measure how children are exposed to chemicals in their environments. But Leckie, the C. L. Peck, Class of 1906 Professor in the School of Engineering, may have achieved his greatest success when he decided to study children in the first place.
"When we began this work, most of the regulations were based on studies of adult white men -- the healthiest segment of the population," Leckie will tell an audience Feb. 16 at a symposium on toxic substances in the environment at the annual meeting of the American Association for the Advancement of Science.
The foundation of Leckie's research is careful observation of the activities of normal, healthy children in everyday environments. His video cameras have captured hours of mundane activity in the homes of middle-class suburbanites and migrant farm workers.
It is not yet clear whether trace amounts of synthetic chemicals, increasingly present in the environment, pose a risk to human health. But it makes sense to begin to consider the issue by studying children, the most vulnerable segment of the population.
Compared to adults, young children have cell walls that are more permeable to environmental toxins. Babies are born with incomplete liver and kidney functions; the ability to detoxify chemicals matures with age. Children also have higher surface-to-volume ratios. So if a child and an adult are exposed to the same environment and absorb toxins through the skin, the child will receive a greater relative dose.
Leckie has used the hours of footage to build realistic models of how children might ingest or otherwise become exposed to substances, including toxins such as lead and pesticides, which may be present in their immediate environments in small amounts. The development of these models has changed significantly during the last 10 years, in part because of Leckie's work.
"In the old days, researchers were limited by observation tools at their disposal," Leckie said. "They would watch children in real time armed merely with a stopwatch and a pad of paper."
This approach might yield as many as a few hundred data points per observing session, but much was inevitably missed. And in the days when graduate students were cheap and computers were expensive, these incomplete data were expensive to process.
Today things are different. With funding from the Environmental Protection Agency (EPA), the Research Triangle Institute and the National Institute of Environmental Health Sciences, Leckie has devised an approach that allows for the collection of tens of thousands of data points in a similar session.
Does the child onscreen touch a toy with his right hand while playing and then, while resting a few minutes later, rub his eyes with that same hand? Does his left hand touch two different toys in the interim? By watching the videotapes over and over again, concentrating on a different body part each time, the students Leckie hires to collect data eventually capture a holistic picture of exposure risks by making several entries about the same child into a database.
Custom software is used to record information on up to 16 body parts, 36 object types, three different activity levels and 12 microenvironments. These data, combined with today's cheap, powerful computers, give a detailed analysis of a child's interaction with his immediate surroundings.
Valerie Zartarian is an environmental engineer with the EPA. She studied with Leckie as a doctoral student at Stanford in the mid-1990s when he was beginning his work on this topic. For her dissertation, she developed a model to quantify exposure to pesticides resulting from hand-to-surface and hand-to-mouth contacts among a group of videotaped children. Zartarian's real innovation, inspired by her work with Leckie, was to take a "microlevel," or second-by-second, look at the children's activity.
It's an innovation that attracted the attention of her current employer. EPA's models for assessing children's overall exposure to toxins from multiple routes -- inhalation, ingestion, dermal, hand-mouth, object-mouth ? rely in part on Zartarian's time series work. So does the agency's recent attempt to determine if arsenic-treated wood increases children's cancer risk. Arsenic is used to treat wood in play sets and decks because it repels wood-boring insects.
Trace amounts of everything from ibuprofen to DDT are showing up in ecosystems around the world. Some environmentalists suspect these and other synthetic compounds in behavior changes and increased rates of mutation among fish and amphibians. Yet not much is known about the potential risk to human health.
Richard Beck, a Ph.D. toxicologist with the American Chemistry Council, an industry trade association, said that little is understood about the rate at which chemicals are absorbed or ingested into the body, so the work of university researchers like Leckie to calculate exposure is important. "There is a real absence of good science in this area," said Beck.
But Beck cautioned against being overly concerned about exposure to trace amounts of chemicals, and pointed to ethanol as an example. Exposure to too much ethanol is associated with neurological, reproductive and liver problems. Yet ethanol, produced in massive quantities as a fuel additive, is also a naturally occurring metabolite that is common in many foods, including bread, in tiny amounts.
"You always have to consider the dose, not just the compound itself," Beck said.
Leckie is no alarmist. He regularly eats microwave popcorn even though he knows he is ingesting small amounts of perfluorinated hydrocarbons, synthetic chemicals that repel water and oil. No one knows if the compound, used to line microwaveable bags, poses any risk. For now Leckie has no plans to forgo one of his favorite snacks.
"The human animal is pretty tough," Leckie said. "But the problem is that we're pretty ingenious, too, when it comes to designing synthetic compounds that last in the environment."
Leckie has applied his own ingenuity in a somewhat novel way for someone in his discipline. Often environmental engineers work on cleanup projects. Over the years Leckie has had to learn much in the way of human behavior and physiology to be successful in his research. He hopes that this somewhat wide-ranging approach might someday inform legislation that will help to prevent future health risks to children.
Many laws to protect the environment and human health are already on the books. The Food Quality Protection Act, passed by Congress in 1996, recognized the importance of looking at children as a risk group. The law requires the EPA to address risks to infants and children before new pesticides are introduced and to collect better data on food consumption patterns, pesticide residue levels and pesticide use.
It's good legislation, Leckie said, but the challenge is to get politicians to realize that even the best models today are works in progress.
Leckie now hires undergraduates -- cheaper than graduate students but still pricier than computers -- to analyze the videotapes. With dozens of body parts to track second-by-second, it's time-consuming work.
But it's important work, too. The better the data, the more likely the models are to paint an accurate picture of the exposure to potential toxins in the environment.
"As good as our statistical and computation tools are, we still need a robust dataset," Leckie said.
Geoff Koch is a Stanford graduate student in journalism.
Leckie will participate in the symposium "Toxic Substances in the Environment: Geochemistry and Health" at the annual meeting of the American Association for the Advancement of Science in Seattle on Monday, Feb. 16, from 10 a.m. to 11:30 a.m. PST. A photo of Leckie is available on the web at http://newsphotos.stanford.edu (slug: "AAAS"). This release was written by Geoff Koch, a Stanford graduate student in journalism.