Some babies are just in a hurry. A walk into any hospital’s neonatal intensive-care unit confirms it. Last year, one-tenth of all births in the United States came before the pregnancy’s 37th week, premature by current standards. The severest cases wind up in the Nick-You, as the residents pronounce it—a clamorous, liminal place. Beeps, bells, nurses’ chatter, endless crying. Babies in translucent boxes, fed sugar through plastic intravenous tubes, their lungs suffused with surfactants to prevent collapse. Many of the infants survive; prematurity is no death sentence. But it can bring a complicated life: blindness, cerebral palsy, autism.
The causes of prematurity remain mysterious. There are known risk factors: smoking, poor prenatal care, in vitro fertilization. It is a disorder of modern life; the U.S. rate peaked only eight years ago. Mississippi is a serial offender. Even worse is Puerto Rico, where nearly one-fifth of the babies are born early. Worldwide, only Malawi has a consistently worse preterm-birth rate than Puerto Rico. No one knows why.
Until he returned to his native Puerto Rico, José F. Cordero had not given much thought to preterm birth. Cordero, a physician and veteran of the federal Centers for Disease Control and Prevention, had spent much of his career investigating birth defects. But after three decades away, he got a call: Come home and lead the University of Puerto Rico’s Graduate School of Public Health. Once he arrived, in 2006, he saw a crisis. He had to understand, and try to stop, the island’s surge of preterm births.
Ricardo Arduengo for the Chronicle Review
Ingrid Padilla, a hydrologist, runs a test in Arecibo. “People don’t really know the consequences of the contamination” from drug manufacturing, she said. “And some people don’t want to know.”
Soon after Cordero’s return, two engineers appeared at his door. One, Ingrid Y. Padilla, was a slight, wry hydrologist from a west-coast campus of the university; the other, Akram N. Alshawabkeh, an expert in groundwater remediation, came from Northeastern University. They told Cordero that Puerto Rico has one of the nation’s highest concentrations of Superfund sites, those peak indicators of contamination. Much of the pollution is found on the north coast, west of San Juan, where decades ago pharmaceutical companies, spurred by tax incentives, set up shop. After those tax breaks expired, in 2006, the manufacturers fled, but their waste lingered.
The engineers had their eyes on a Superfund-focused research grant at the National Institutes of Health. Surely this chemical waste must pose a health risk, but they were at a loss to say what, or how. Cordero had just the problem for them. His team had looked at the risk factors for preterm birth, and none explained Puerto Rico’s high rate. Could it be something in the water?
Cordero knew that it would be difficult to link preterm birth to pollution, especially at low doses. Biology rarely grants simple explanations. Except in the case of the most obvious pollutants, fierce battles rage over how proven toxic effects in animals, often at high doses, can be extrapolated to humans.
Everyday chemicals might lead to premature births.
They decided to court the controversy. “We didn’t want to do the typical contamination,” Alshawabkeh says, pollutants like mercury or arsenic. Instead they’d investigate phthalates, common plastic additives that, like their relative bisphenol A, have raised consumer hackles for a decade over fears that, even at low doses, they can disrupt hormone function in humans. They brought in phthalate experts from the University of Michigan and cited their preliminary evidence showing elevated levels tied to a handful of Mexican preterm births. NIH gave them $9.9-million in 2010 to support four years of work. The researchers dubbed the project “Protect.”
Phthalates (the “ph” is silent) are nothing if not elusive. Most commonly they are mixed among rigid chains of carbon in PVC, serving as molecular grease; without them, there would be little “plastic” about many plastics. They come in varieties heavy and light, united by a common chemistry: two acetate wings waving from a ring of carbon atoms. The heavies reside in building pipes, flooring, toys; the lighter molecules in food containers and skin creams. Since they live in plastic but are not bound to it, phthalates are notorious for leaching into the human body. But they are not long-term guests: Our bodies chew through phthalates; levels are kept high through sustained exposure. Children, never afraid to lead explorations with their mouths, are prone to high phthalate levels. That’s one reason that, despite limited evidence, Congress restricted the use of several phthalates in toys six years ago.
One morning this winter in Puerto Rico, I was driving west from San Juan with Ingrid Padilla, searching for the Zanjas Frias, a natural spring she uses for groundwater sampling. Padilla herself had been a preemie, had gone on to become an NCAA diver, and was now an expert in the snakes-and-ladders hydrogeology of her homeland. Water inspires her. At one point we walked into a cave, and she pointed toward an inchwide bore in the floor, a droplet of water at its base. “The water drop is very persistent,” she said. “Little by little you start making that hole, and eventually it becomes very large. Being persistent. I always say you have to be like water.”
As we drove west, the highway sashayed through abrupt hills, mogotes, crowned with African tulip trees; the hills’ egg-box topography was a hint of the karst limestone system below us. Formed by the fossilized remnants of coral reefs, karst defies easy analysis. As rain falls on Puerto Rico’s mountainous east-west spine, it is shunted underground through tunnels eroded into the limestone. Groundwater does not percolate here; it flushes. Invisible rivers run underfoot, periodically exposed by the tree-swallowing arrival of a sinkhole.
Several of the island’s more than 200 contaminated sites loomed. There was the boxy white facade of the Upjohn Facility, a shuttered pharmaceutical plant, where a tank of carbon tetrachloride leaked in 1982. Not far away was the Barceloneta landfill, an unlined dump built into three sinkholes. Both sites are Superfund-listed; their pollution is considered contained. Or mostly contained: Padilla has seen early signs that the karst may allow pollution to build up in areas far removed from their defined boundaries.
These were marshlands, the karst’s freshwater shoved upward by dense ocean water. We were near Garrochales, a poorer part of the island. Satellite dishes pockmarked pastel-concrete houses. Cows grazed. Many of the local farms have groundwater wells, but the owners are loath to let Padilla sample them. “People don’t really know the consequences of the contamination,” she said. “And some people don’t want to know.”
We were also lost.
It had been some time since Padilla visited the Zanjas Frias. We drove north, toward the coast and past another landfill, then doubled back, catching a right at a black Everlast punching bag hanging under a monkeypod tree. Figures appeared down the dirt road—three teenage boys on horseback. We were close to Zanjas Frias, one told her, but the road was overgrown. We’d have to hike in. Don’t leave the car for long, he added: “There’s a lot of weird people that can come by.”
Early on, Padilla faced doubts that she could even detect phthalates in the groundwater, given their propensity for breaking down. That hasn’t been the case. The researchers have found them, though only in the parts-per-billion range.
Those field numbers weren’t a perfect match for what the researchers were finding in the tap water of the mothers, where more than half of the samples had low levels of phthalates. “You’re increasing the detection, meaning there are other sources,” Padilla said. “They’re not just coming from contaminated sites.” She hopes to expand her collections, and her students are developing models of how phthalates might escape from landfills.
We bushwhacked in, hiking past a stagnant, algae-clogged marsh; these pools weren’t clear to the bottom, like her spring. We were lost again. Overgrowth blocked our view back. Padilla stopped. “I’m concerned about the car,” she said. Zanjas Frias would remain elusive.
Even in ideal circumstances, epidemiology is tricky, prone to false associations. And premature birth is far from the ideal subject: Even in Puerto Rico, most study participants will not have premature babies. That’s why the team had to recruit a fleet of expectant mothers—some 2,000 women across the north coast, amassing 2,500 data points on each one. They would all have their particular tribulations. It was up to Liza V. Anzalota Del Toro, Cordero’s protégé, to gather them.
Ricardo Arduengo for the Chronicle Review
Liza Anzalota Del Toro (right) recruits participants for a continuing study of groundwater contamination and premature births in Puerto Rico.
Anzalota is the type who, after a long work week, considers her weekend job as an emergency-room doctor good therapy. She splits her time between Cordero’s office, where she’s an adjunct professor, and nightly shifts at a local clinic. She takes just two days off a month. Her nurses still joke about when she took Alshawabkeh to visit the clinics, and he was too polite to note that they hadn’t stopped to eat all day: “You’re starving Akram!” they yelled.
I joined Anzalota for several days of clinic visits. Our first stop was a private OB/GYN in Manatí, 30 miles west of San Juan. Its doctor takes pride in being the top contributor to the Protect project. He’ll even haggle with a patient if she decides not to donate her umbilical-cord blood or placenta to the study—not strictly protocol, but hey, it happens. Anzalota has done many participant screenings herself in the single-story building’s white-tiled supply room, lined with boxes of baby bottles and breast pumps.
Today there was one expectant mother who might qualify for the study. Anzalota ushered her into the supply room, along with her husband; including family members is a tactic to improve retention. The couple were in their early 20s, and the wife, with dyed-blond hair and a tie-dye top, sat at the edge of her seat, resting chin on fist. The husband, in black T-shirt with a fanny pack thrown around his shoulder, leaned back, stealing glances at his phone.
Anzalota introduced the study and some exclusions: The mother had to be 18 to 40 years old and less than 20 weeks pregnant. There would be three visits, she told them, to collect blood and urine samples. There would be questions about the mother’s history, everyday life, stress, and support network. The second visit would come at home, and include tap-water sampling and a questionnaire on food habits and personal-care products. One more visit would come after the baby’s birth. And they’d get $50 for their time.
With an “OK,” the beaming wife ended the 20-minute interview, and her husband gratefully fled the room. She would be Anzalota’s only recruit of the day. It was a slow process. “From February to April, that’s our peak of recruitment and follow-up,” she said, “because, of course, during December people get festive, and a lot of women get pregnant.”
The Manatí clinic has provided more than 400 participants to Protect—so many that it threatened to skew the sample. Its patients tend to be older and wealthier, so when Anzalota started, she had to lure several publicly financed clinics into the study. That was one of many early kinks.
Later we were driving into the foothills, Anzalota juggling a phone and coffee, using her knees to swerve the SUV around an orange-vested street cleaner. The best participants, she said, become citizen scientists, sampling their own tap water, investigating local pollution, recruiting cousins. It gives them a voice, one nurse told me.
We pulled into another clinic, this one publicly funded. Abril Burgos, a young mother who has had three premature babies, was waiting for us. A bad case of reflux had struck her first baby, and until her participation in Protect, it had never occurred to her that there might be an environmental influence.
The research project may not have results yet, but it’s given Burgos a cause. She recruits local women and wants a career in neonatal care. At her last delivery, between gasps, she even had a stern reminder for her doctor: “Make sure you collect the placenta and cord blood,” she said.
If there’s a missing connection between Padilla’s hydrology and Anzalota’s epidemiology, it’s biology. How in the world could phthalates, such everyday chemicals, be involved in preterm birth?
Ricardo Arduengo for the Chronicle Review
José Cordero, dean of public health at the U. of Puerto Rico, knew it would be difficult to link high rates of preterm births to pollution. Biology rarely grants simple explanations.
The team recently bolstered its hypothesis with a study in Boston, published late last year. Based on 500 previously collected samples, the study found a link between several phthalates—especially di-2-ethyl hexyl phthalate, a primary chemical of concern—and preterm birth. But it’s just a correlation, said John D. Meeker, an associate professor at the University of Michigan at Ann Arbor who leads Protect’s epidemiological analysis. Diet could be the cause; the women could have eaten processed food that exposed them to phthalates. This a frequent problem in epidemiology, and the reason that Protect’s women are being questioned so thoroughly about their eating habits.
The Boston study was welcome news this winter, when Protect team members met for their annual meeting, at a spring-break-frayed beach resort east of San Juan. Circled around a dim conference room, the scientists prickled. Their grant was up for renewal at a time of budget-tightening, and they needed to produce results. They had recently received the peer reviews for their application and were working through them. Coffee flowed, tensions relieved by strolls out in the sun. The scientists groused that perhaps the reviewers had read the wrong application, because it was all in there already.
Rita Loch-Caruso was having the roughest time. A toxicologist at Michigan, she examines how phthalates could contribute to preterm birth. Researchers have long pointed to uterine inflammation as a possible culprit, and she had an idea that some phthalates could cause stress and, in turn, inflammation. Or, as she put it, how might they “treat HTR-8 cells with MEHP, assessing ROS generation, oxidative DNA damage, redox-sensitive gene expression, and apoptotic cell death?”
It wasn’t easy for Protect’s researchers to find a common language. You should have seen the early joint webinars, Anzalota said. “Many of us, including faculty, we were really lost. You have to translate the field epidemiology to an engineer’s perspective, and the field epidemiology from the public-health perspective. The analysis of biological samples from the toxicologist’s environment. And we were all very focused on what we did.”
When it came to bridging disciplines, molecular biochemistry seemed like an ocean. But they found passage points. One came early between Loch-Caruso and Alshawabkeh, the engineer, who was creating a solar-powered device to clean up the pollution. His machine uses electricity to radicalize oxygen in the groundwater; primed for reaction, the radicals dismantle pollutants. While Alshawabkeh described his work, Loch-Caruso realized that it was similar to a reaction used by the body to create energy. They were working with the same tools, just in very different contexts.
Loch-Caruso has several research lines open. Parallel to phthalates, Protect has looked at trichloroethylene, or TCE, a classic pollutant. Their dual targets were an important part of winning the grant. But Loch-Caruso was not having much success with TCE. “Should I throw TCE out?” It was a question she asked on the first day of the meeting, and one that no one particularly wanted to hear. If they wanted to win a grant extension, they had to keep it in. There were ways to make this work, as one adviser told Loch-Caruso: “Say you’re going to develop the ideal model for testing these things.”
Loch-Caruso has done a lot of work with rats, but the animals were never going to be a well-accepted analogue for preterm birth, a member of the scientific board said at the meeting. Their gestation time is driven by litter size, “which makes it less relevant when you’re thinking about human pregnancy,” he explained. Maybe she should use monkeys instead. Loch-Caruso was wary: “I’d need a veterinarian who knows how to do C-sections on rhesus monkeys.”
She’s had more luck with human tissue. Working with her hospital’s operating-room staff, her team members dissect protective membranes that, minutes earlier, were encasing infants in the womb. Lying on the lab bench, the membranes resemble bloody crepes. The researchers excise a stamp-sized slice and suspend it between liquids using what looks like a shot glass. They have evidence of stress signals’ traveling across the membranes, and, by exposing the skin to a derivative of diethylhexyl phthalate, the researchers seemed to cause an increase in oxidative stress and the expression of prostaglandin, a molecule known to trigger birth.
The reviewers, however, were unhappy with the doses Loch-Caruso was using, which were 10 times as high as what you might see in the real world. “They want me to be exposing my in vitro cultures to concentrations that are relevant to Puerto Rico,” she said. “Now, the fact that I didn’t know what those concentrations were until very recently was irrelevant to the reviewers.” She gave a weary laugh and pointed to a fundamental problem: She didn’t know how to extrapolate the urine biomarkers collected from Protect’s participants to her own lab system. Translating doses is an endlessly complicated problem.
“Sorry,” she said. “I could add somebody to my project who could do it. But I’d need more money.” The reviewers moved on.
Ambitious science, by definition, carries a risk of coming up empty-handed. The Protect research, with its extended hypothesis, is riskier than most. When I visited, its scientists were confronting the reality of results against the dreams of ideas. I was reminded of decoherence, to borrow a term from quantum computing: adding more bits to the system increases the likelihood of collapse.
Everyone had conceded something or had had a link fizzle. It would be easier for Alshawabkeh to test his device in a more traditional aquifer, not in karst, with its fast-moving fluids. Padilla’s groundwater tests did not easily match with data from the study’s participants—some have wells, and others municipal water. Either way, contamination may not be the most important phthalate source. The Superfund sites were a source of the pollutant TCE, but for preterm birth, TCE may not matter. And while the tests had shown that a derivative of di-2-ethyl hexyl phthalate in the Protect women was twice as high as in the U.S. general population, they were far from establishing a genuine tie.
People like easy answers, but rarely do chemicals work in isolation. Take the bisphenol A out of the toys, and bisphenol S remains. Is that a problem? It’s usually a mistake to think that a single chemical could have a clear effect on a complex health problem. Protect needs to hedge its bets: As part of its second phase, it would try to catch other important tracers in its samples.
“It would be great if we were able to identify a smoking gun, because then some sort of intervention could follow quickly,” said Meeker, the Michigan epidemiologist, a few months after my visit. “But I’m trying to be realistic in my expectations.”
Later that month, the team received good news: NIH had granted it an additional five years. The project was guaranteed to reach full term.
My second day with Anzalota, Cordero’s protégé, we drove to Ciales, a mountain coffee-growing town. She hoped to stay with Protect for at least another year before pursuing a residency. She’ll leave knowing that engineers, doctors, epidemiologists, and toxicologists can find common ground. “We’re now speaking the same language,” she said, a point that NIH commended them on in their review. “We’re speaking the language of public health.”
We stopped at the Ciales clinic, another publicly financed venture. After a tour and more screenings, Anzalota noticed a woman wielding a bulky stroller in the overflowing waiting room: a Protect participant who hadn’t been scheduled to visit but had a cyst she wanted investigated. Anzalota gave her a hug and pulled her aside, commandeering an office. I stood to the side, the woman’s month-old daughter swaddled beneath me in a sea of pink fleece, a polka-dot bow on her head. She clenched one miniature fist, and her eyes fluttered.
It had been a vaginal birth, the baby arriving at 39 weeks, weighing six pounds, the mother said. Not preterm.
“It was very quick? Yes?” Anzalota said.
“Yes.” Only about three hours.
“Nothing! You weren’t in it at all.” She gasped. “What a child!”
A Ciales resident, the mother believes that she lives in contaminated land. She had been worried—her brother’s ex-wife had had a preemie who died during a bungled delivery. And she herself had risk factors, as everyone does—maybe even high phthalate levels. But her baby was healthy. Anzalota looked forward to Facebook photos. They had another data point. They’re up to more than 700 women now. Only 1,300 more to go.
Wide awake now, the newborn looked up and sneezed.
Paul Voosen is a senior reporter for The Chronicle.
