Photo courtesy Guido Vitti

"We need the environmental side becasue the Human Genome Project isn't going to us all the answers," Perera says.

On an early spring day in New York City, a clean wind from the north sweeps down the Hudson River. Cars are backed up on the George Washington Bridge, their tailpipes spewing, yet the air today seems to brush the pollution away. It is so clear I can make out every fissure in the rust-colored cliffs of the Palisades across the river in New Jersey. What a terrific view Frederica (Ricky) Perera has from her 25th-floor office.


However clear the day, the view is deceiving. For 25 years the 64-year-old professor of environmental health sciences at the Columbia University Mailman School of Public Health has been probing the long-term, invisible impacts of air pollution on health. An award-winning cancer investigator and defender of the welfare of newborn children, Perera comes from the public-health model of disease, which assumes that most ailments are conveyed from outside the body and can be prevented. She has pioneered a field called molecular epidemiology, a hybrid science that melds urban surveys with subtle molecular changes. Her work ranges from the noxious tailpipe to the precancerous cell, evaluating all the possible way stations of disease. It is an extremely complicated task because it is so broad. Progress in molecular epidemiology has been slow, but Perera is not one who gets discouraged.

Just blocks from her base at the Columbia Center for Children's Environmental Health are the low-income neighborhoods of Washington Heights and Harlem. The poor there tend to live with more pollution than other people do. Some is of their own making, like cigarette smoke, but a lot of it they cannot avoid, like lead in old paint and smoggy urban air. The predominantly African American and Dominican subjects of her research live a world apart from Ricky Perera, yet she thinks about their health all the time.

Since it began in 1998, her Mothers and Newborns Study has enrolled 700 women. The project monitors women's exposures to airborne chemicals during pregnancies and tests their babies as soon as they are born. Tracking particles of pollution that pass from mother to child, Perera and her team have connected the process to lower birth weights and smaller head circumferences in some infants. She suspects cancer could be an outcome as well, although it's too early in the study to know for certain.

Perera has agreed to take me to a clinic where participants in her studies are recruited. Wearing black slacks and pale makeup, she puts on a black leather jacket and a black leather backpack. Thin and athletic, she walks at a rapid clip down 168th Street. When we get to the Audubon Clinic, which serves low-income patients and is supported by the university, we sit in the corner of the waiting room, trying to be unobtrusive. Perera's eyes flick about for pregnant women.

An assistant with a bunch of flyers strolls in front of the young women waiting in plastic chairs. Because of new rules protecting patients' privacy, the staffer cannot give them a hard sell about joining the study. Rather, she simply asks women if they would like some information about a research project. It helps that each mother-to-be in Perera's study receives a series of small payments.

The first research step, she says, is "collecting dust and air samples and interviewing the mom at home." After the pollutants are recorded, the next step is to look for biological signs of chemical exposure, which she calls markers. Some markers may represent early fingerprints of disease.

The simplest markers show concentrations of foreign substances in blood or fat. Take lead, perhaps the most dangerous of common pollutants. The amount of the metal in a child's blood has proved to be a reliable indicator of the amount of neurological or cognitive damage following exposure to leaded gasoline or paint chips. Although a mother's placenta is usually a barrier against many unwanted chemicals, lead, like some other chemicals, passes directly from the mother to the fetus. As Perera notes sardonically, "One way to get rid of lead is to have a baby."

Her favorite chemicals—favorite in the sense that she has studied them more than any others—are the polycyclic aromatic hydrocarbons in cigarette smoke, power-plant emissions, automobile exhaust, and other sources of combustion. These compounds cause cancer in laboratory animals, and studies of industrial workers strongly suggest they can cause lung cancer in humans too. Near the end of her pregnancy, each woman wears an air monitor, a small pump-and-filter system that records the hydrocarbons she breathes over 48 hours. No smokers are enrolled in the research, but many of the women report they are exposed to secondhand smoke at home or at work.

When a participant goes into labor, she is supposed to notify the Columbia center. A staffer retrieves the placenta and draws a tube of blood from the umbilical cord, in effect taking a sample from the baby. If possible, blood is collected from the mother, too, and the lab later identifies and compares markers in the samples. Perera isn't interested in raw amounts of pollutants in the subjects' tissue or fluids. Instead, she looks for signs of hydrocarbon exposure in the genetic material of the white blood cells, because chemical interference with DNA can initiate cancer.

The National Institute of Environmental Health Sciences, the health agency that pursues the environmental causes of cancer, has been a major funder of Perera's center. The institute has seen its support pay off in 54 publications by Perera's group in the past four years. A paper published just before my visit has caused a stir of media interest. The Columbia investigators analyzed three possible links in the tortuous chain from chemicals outside the mother to cancer inside the child.

One reading measured the levels of harmful hydrocarbons in the air that 60 mothers inhaled in the weeks before they delivered. The other two measurements recorded levels of two markers in white blood cells of the women and their infants. One is a type of molecule, called an adduct, in which a hydrocarbon locks onto a part of the DNA. The other marker is an aberrant pattern in the chromosomes; the DNA shows a greater number of abnormalities than normal, including DNA that is missing or moved. According to other research, both classes of markers—the adducts and the chromosomal aberrations—are associated with increased risks of all cancers.

Perera's paper in Cancer Epidemiology Biomarkers and Prevention linked the chromosomal aberrations, which were somewhat higher than normal, to air pollution: On average, the higher the hydrocarbons in the mothers' air, the more frequent the abnormalities seen in the infants' chromosomes. The research did not claim a cause-and-effect relationship between hydrocarbon pollution and an indicator for cancer. But after the institute and the Columbia group put out press releases addressing the obvious implications—Perera edits her center's releases line by line—the local newspapers connected the dots. "Bad City Air Boosts Kids' Risk of Cancer" and "Cancer Is 'Air' Born" were two of the headlines.

 

In a radio interview Perera cautions that the results do not necessarily mean a child will get cancer if the mother breathes polluted air. She says it isn't practical for women to think of moving to the country to escape air pollution. Rather, she says, policymakers should review pollution standards to see if the regulations are protective enough. I notice that Perera likes to throw this one-two punch—her research points directly to problems she thinks society should resolve.

In her office I question her about a finding the news media had ignored. The DNA adducts in her study weren't linked with the other two factors that she measured. This weakened the results, because three measurements lining up in a sequence to cancer are a lot stronger than two.

She is not fazed. "Contrary to our hypothesis," she replies cheerfully, "the exposures correlated stronger [to the chromosomal aberrations] than the adducts did. In fact, the adducts didn't correlate at all. Maybe it was due to the small sample size."

Another puzzle is that the African American newborns showed substantially more aberrations in their blood than the Dominican babies did, given the same range of exposures. Why would the two groups be so different in their responses to low levels of pollution? That definitely needs investigating, she says.

But by now she has picked up the implication of my questions: Do her social concerns taint her research results; is she biased? "I don't like advocacy without a scientific basis," she says firmly. "The passion without the facts won't work."