Kim's Tuesday morning visit to the doctor on May 27, 1997, was supposed to be no big deal, just a regular checkup to see how everything was going. At the time, we were 24 weeks pregnant with twin girls.

I was working in my home office that morning when Kim called from her doctor's office in Winchester, Va.

"They're sending me to the hospital," Kim told me matter-of-factly. She said her blood pressure was high and that she had protein in her urine -- two signs of a pregnancy-related disorder called preeclampsia. Kim said it probably meant she'd have to spend the summer on bed rest. They were planning to do more tests at the hospital.

When I asked how she was feeling, Kim said she was fine. "I feel perfectly fine," she said. "I don't understand it."

But Kim wasn't fine. After tests showed she had a severe form of preeclampsia and her life was in danger, she was sent by ambulance from the hospital in Winchester to the University of Virginia Medical Center in Charlottesville. Doctors told us our girls would have to be delivered in a matter of days, maybe hours, to save Kim's life. At UVA, with its high-level newborn nursery, the babies would at least have a better chance to survive.

Extracted from their mother two days later--nearly 16 weeks before they were due--the girls weighed 1 pound 6 ounces and 1 pound 2 ounces and immediately were put on ventilators and sent to the newborn intensive care unit. The smaller one, Nina, died 49 hours later after a valiant struggle for life. The other, Josie, now more than 2 years old, is an active toddler who delights in walking our dogs and eating grilled-cheese sandwiches. Although she is still a few months behind in her development, the doctors say it won't even be an issue by the time she is in kindergarten.

After the delivery, Kim was over the preeclampsia symptoms in a matter of days.

A Medical Mystery

A mystifying condition that can lead to seizures and death, preeclampsia affects as many as one in 10 pregnancies in the United States. Researchers still are trying to understand what triggers it. The primary signs of the disorder are high blood pressure, protein in the urine and swelling, or "edema." Known in the past by a variety of names from "toxemia" to "pregnancy-induced hypertension," preeclampsia is the leading cause of maternal mortality in the United States and accounts for 75,000 maternal deaths each year worldwide. Babies born to preeclamptic women face a significantly higher risk of dying or developing lifelong disabilities than babies born to healthy mothers, primarily because the only known "cure" is delivery of the fetus. The condition accounts for one-third of all premature births, although many women bring the fetus to term with bed rest and medical treatment.

James M. Roberts, an obstetrician-gynecologist at the Magee-Women's Research Institute in Pittsburgh, has spent more than 20 years researching preeclampsia. He heads a team of 20 physicians, PhD's and postdoctoral researchers who comprise the nation's largest program to investigate the condition.

Roberts said he first became interested in preeclampsia while working on a postdoctoral fellowship at the University of California, San Francisco. After he completed his residency, he wanted to learn more about hypertension. And as he encountered significant numbers of pregnant women suffering from high blood pressure and preeclampsia's other effects, Roberts became convinced that this was something different--that hypertension was the least of these women's problems.

"It was so amazing to see how tremendously ill these women had become and what a major impact it had on them and their babies," he said. "And yet when they'd come back six weeks after delivery, they were right back to normal. It was just amazing from a physiological perspective--really bizarre."

In a 1989 article in the the American Journal of Obstetrics and Gynecology, Roberts and other researchers first encouraged the medical profession to dismiss the notion that high blood pressure was preeclampsia's defining feature and to focus instead on other changes occurring in the patient's body.

Specifically, Roberts and his colleagues suggested that injury to the endothelial cells that line blood vessels and promote healthy blood flow might play a starring role in the onset of the condition. They went on to propose that endothelial cell injury in preeclamptic women was caused by something unknown circulating in their blood--a substance that the researchers theorized was released by the placenta. Preeclampsia had been linked in numerous studies with reduced blood flow to the placenta and other organs. Roberts and his colleagues wondered if perhaps the placenta was responding to the problem by releasing something dangerous into the woman's blood.

A Mishmash of Theories

It was not until the invention of the blood pressure cuff in the 1890s that doctors established a link between pre-eclampsia and elevated blood pressure. The very idea of "pre-eclampsia" is relatively recent. Until the mid-1800s, eclampsia, or the seizures that follow preeclampsia, was considered a pregnancy-related disorder that happened out of the blue. Hence the use of the term "eclampsia," from the Greek for lightning.

As doctors studied the condition more closely, they realized that preeclampsia was a harmful condition in and of itself, even if the woman did not go on to have seizures. The primary problem is "poor perfusion," or reduced blood flow to the placenta and other vital organs. For the fetus, this can mean restricted growth; more than one-third of the babies of preeclamptic women are born small for their gestational age. For the mother, poor perfusion can cause damage to organs such as the liver and kidneys, among other problems.

Over the years, researchers have come up with a mishmash of theories and counter-theories about what causes preeclampsia. "It's almost funny if you look back at the history and see all the different ideas," said Roberts. "People have thought that too much protein was bad, too little protein was bad, too much fat was bad, too little fat was bad, too much salt was bad, too little salt was bad. Take your pick."

In recent years, it was suggested that taking low doses of aspirin early in pregnancy might help prevent preeclampsia. The aspirin theory was based on several small studies that had shown a drop in the incidence of preeclampsia among women at high risk for the condition who were given the drug. Larger studies conducted under the aegis of the National Institutes of Health (NIH), however, threw the earlier conclusions into doubt.

The latest target of pre-eclampsia research was calcium. Epidemiological studies of women on low-calcium diets in Central and South America in the 1980s suggested that they were significantly more likely to develop preeclampsia. Again, however, a subsequent study by the NIH found "absolutely no effect" of increased calcium intake on the incidence of the disease.

Yet despite the years of false starts, the past decade has seen enormous progress in researchers' understanding of pre-eclampsia.

Roberts and others now believe that the condition can be traced to events occurring very early in pregnancy. In a normal pregnancy, the maternal blood vessels to the area of the uterus where the pancake-like placenta attaches to the uterine wall undergo a stunning transformation. To supply the uterus and placenta with a sufficient flow of blood to support the growing fetus, these blood vessels, called the "spiral arteries," increase in diameter by a factor of four, the equivalent of a garden hose assuming the girth of a 24-ounce can of Foster's Lager. These changes are accompanied by an equally dramatic increase in the mother's total blood volume, which can expand by as much as 50 percent during the course of a pregnancy; two-thirds of the increase in blood volume happens by eight weeks gestation.

"It's just boom, things explode," Roberts said. "And all these changes seem to be directed at preparing the woman to carry this big mass of baby with a huge placenta that she has to perfuse."

It was recognized in the 1960s, however, that in pre-eclamptic women the blood vessels to the uterus generally don't undergo the same transformation. The spiral arteries look much like they do when the woman isn't pregnant, a clear indication that the placenta isn't getting an adequate supply of blood.

Why the spiral arteries in preeclamptic women remain small enough to inhibit the flow of blood and oxygen to the placenta is a question at the heart of Susan Fisher's research at the University of California, San Francisco. Fisher, a cell biologist, explained in an interview that the growth of the placenta in the early weeks of pregnancy is a result of the rapid division of placental cells called cytotrophoblasts. After 10 to 12 weeks, the cytotrophoblasts stop dividing and suddenly are transformed into wildly invasive cellular foot soldiers that penetrate the uterus and maternal blood vessels in the same way that tumors attack tissue.

During a process called "endovascular invasion," the invading cells are capable of mimicking the function of the endothelial cells lining the mother's blood vessels--a trait that earned the cytotrophoblasts the title "masters of disguise" in a recent medical journal headline. These "differentiated" placental cells thus replace the internal lining of the mother's blood vessels, anchoring the placenta to the uterus and causing the spiral arteries to expand in diameter so they can bring increasing volumes of the mother's blood to the placenta. The process by which the placenta attaches itself to the uterine wall is called "placentation."

"It's an incredibly wild process, and it's driven by these very unusual, tumor-like cells that would kill you if they stayed in this invasive state for a longer time," explained Fisher.

In findings reported last year, Fisher and her colleagues noted that in preeclampsia the placental cells never fully make the transition into an invasive mode. Working with placental tissue cultures from preeclamptic and normal pregnancies, the researchers found that in preeclampsia the cytotrophoblasts fail to penetrate deeply into the wall of the uterus. The same cells also fail to take on characteristics of the mother's vascular endothelium, making them powerless to enact the changes needed to boost blood flow to the placenta.

What makes the placental cells in preeclamptic women less capable than their counterparts in a normal pregnancy? In another study released in 1997, Fisher and others reported that cytotrophoblast cells taken from placental tissue samples failed to invade the uterine wall sufficiently when they were exposed to low-oxygen conditions. Fisher also has observed that low oxygen levels inhibit the ability of the placental cells to mimic the properties of the mother's vascular endothelium.

"We believe that preeclampsia is a clear case of placental hypoxia; the placenta isn't getting enough oxygen," Fisher said.

Fisher and her colleagues are now trying to identify the various proteins produced by placental cells under normal and low-oxygen conditions. The hope is that they will find certain "marker" proteins that doctors can look for as a sign that the placentation process has gone awry.

Working along the same line, researchers led by James L. Mills at the National Institute of Child Health and Human Development reported in July in the Journal of the American Medical Association that women who developed pre-eclampsia had a lower level of a chemical called prostacyclin in their blood months before their disorder was diagnosed. Prostacyclin causes the blood vessels to dilate. The researchers suggested that new studies should explore whether giving pregnant women with this blood marker extra prostacyclin is effective in thwarting preeclampsia.

Attacking Foreign Tissue?

It has long been suspected that abnormal placentation problems--along with a variety of other problems in pregnancy--stem from the fact that the placenta and the fetus derive half their genetic material from the father and thus are foreign tissue to the mother. According to this view, problems during pregnancy can be explained as immunological reactions taken to an extreme; the mother's body is essentially trying to ward off an intrauterine invader.

Support for this idea comes from studies showing that the more the mother is exposed to the father's genetic material--either through a prior pregnancy or after several years of intercourse--the less likely she is to develop pre-eclampsia. This could explain why first pregnancies account for as many as two-thirds of all preeclampsia cases. It also could explain why younger women are at a greater risk of developing the condition. There are even studies showing that women who use "barrier" forms of contraception--such as condoms or diaphragms--are more likely to develop preeclampsia; presumably, their system never has a chance to get used to the father's sperm.

David Haig, an associate professor at Harvard University, has attracted considerable attention in recent years with his idea that "evolutionary conflicts" between the mother and fetus can help explain common medical complications during pregnancy.

In the case of preeclampsia, Haig suggested in an interview that the maternal-fetal conflict gets out of hand when the fetus is deprived of an adequate supply of nutrients as a result of abnormal placentation. This, according to Haig's theory, prompts the fetus to "take steps to try to get more out of the mother," regardless of the effect on the mother's health.

An obvious way for the placenta to respond if it is not getting enough blood, Haig explained, is by releasing something into the mother's circulation that will act to increase her blood pressure and, consequently, push more blood and nutrients to the fetus. One way to do this is to cause peripheral vascular damage to the mother--for example, by targeting the mother's endothelial cells for injury.

Fisher disagrees with this theory. "I think it's sort of a flashy way to talk about it but I think it's way too simplistic," she said.

According to Fisher, the theory of genetic conflicts between the mother and the fetus discounts the unique nature of the relationship between the two. After all, she said, the fact that placental cells are allowed to invade the uterine wall and maternal blood vessels without a peep from the mother's immunological system makes it hard to believe there is an inherent conflict at work in pregnancy.

"These cells are from the fetus, from the placenta, and they have only half the genetic complement of the mother. So you would expect to see them targeted by the mother for rapid and violent rejection. But they are not," Fisher said.

The Role of Blood Fats

Roberts's assertion from 1989 that injury to vascular endothelial cells plays an important part in preeclampsia still stands. Numerous studies over the past decade by Roberts and others have shown that blood from preeclamptic women, when applied to a culture of endothelial cells, will change the way they function in dramatic ways. Among other things, the "injured" cells are less able than before to prevent clotting, or coagulation, inside the blood vessel. In addition, the injured endothelium loses its ability to keep the walls of the blood vessels relaxed and loose; instead, the blood vessels tend to stiffen and tighten. This "vasoconstriction," in turn, contributes to the high blood pressure and reduced blood flow associated with preeclampsia.

Clearly, then, there is something in the blood of preeclamptic women that is causing the endothelial cells to change their behavior. But what is it?

As Roberts and his colleagues have studied the problem over the past decade, they've found that a preeclamptic woman's blood differs from the blood of a healthy pregnant woman in many ways--so many, in fact, that it raises an entirely new set of questions about which differences are most important to the progression of the disorder.

Nevertheless, Roberts and his colleagues have focused on several differences that they feel might hold the key to understanding preeclampsia. Specifically, they are looking at components of the blood that they know or suspect can have a deleterious effect on the functioning of vascular endothelium. One of their prime suspects: blood-borne lipids, or fats.

Numerous studies have shown that preeclamptic women have elevated levels of certain lipids in their blood--principally triglycerides and free fatty acids--and that this can be seen before the women become sick, as early as 16 to 20 weeks into a pregnancy.

A piece of the fat puzzle that is of particular interest to Carl Hubel, a researcher on Roberts's staff, is the role in preeclampsia of low-density lipoproteins (LDL), the main cholesterol-carrying proteins in the blood. In a recent study with a group of scientists in Iceland, Hubel was able to show that women who had suffered from eclampsia--the seizures that follow preeclampsia--had abnormal LDL, along with other changes in their lipoproteinsthat could conceivably put them at risk for preeclampsia and then heart disease in later life.

Another study Hubel is working on involves an enzyme that plays a part in clearing triglycerides from the blood. If he were able to find that women with preeclampsia have either low levels or a different kind of this enzyme, it might suggest that these women were more likely to end up with the disorder because of metabolic abnormalities.

Hubel suggested that pregnancy could well be a "stress test" for certain abnormalities in these women that might not otherwise show up until much later in life, if at all.

In a separate stream of research, Hubel and co-researcher Margaret McLaughlin are looking into the role in preeclampsia of free radicals, naturally occurring molecules that are highly reactive and have been shown to cause cell and tissue damage when overproduced. Since Hubel and McLaughlin first suggested in a 1989 journal article that "oxidative stress" caused by the overproduction of free radicals was a key factor in the progression of preeclampsia, numerous studies have affirmed a possible link. In particular, researchers have found that compared with other women who are pregnant, preeclamptic women have lower concentrations of vitamin C in their blood. Vitamin C, Hubel explained, is one of the "front-line antioxidants" that will tend to react quickly in the presence of free radicals.

"That makes it a good marker for oxidative stress," Hubel said. "Whether low levels of [vitamin C] are a consequence of the disease or contribute to the disease we don't know," he added, noting that nagging questions of cause and effect are endemic to the study of free radicals.

Hubel and others caution women against believing that antioxidants will help until larger studies can confirm a possible benefit. Nevertheless, the promise of antioxidant therapy for preeclamptic women--or, more specifically, for women who are identified as being at risk of developing the disease--looms very large over the work of Hubel and his colleagues.

"What you find while you're working on this disease is that you're always discovering new science," said McLaughlin. "It's like every question leads you in new directions, and introduces all kinds of other questions--important questions--that you have to answer first."

"A Whole Different Field"

Far from being daunted by all the questions, the researchers at the Magee-Women's Research Institute seem to thrive on them, if only because their findings continually are adding to our understanding of a very puzzling condition. "I've never gotten discouraged since I started in this area, because I just keep thinking of how much progress we've made in the last 10 years compared to the previous 50 years. It's a whole different field," Roberts said.

On the promise of therapy to help prevent preeclampsia, Roberts confidently added, "There ought to be some way to do it. If you could figure out what was going on--and I think we're getting closer and closer--you ought to be able to make these women better without having to have them deliver."

Cindy Schatzman, a former labor and delivery nurse who is now a member of Roberts's research team, said she can hardly wait for the day when the preeclampsia problem is solved.

"For years I could say we'd never lost a mom here, and two years ago I had to quit saying that. And that's devastating," Schatzman said. "You see the families and what they go through and how it's so frightening, and the hardest thing about it is they have no control. Absolutely none. Everyone is like, 'What did I do? Did I do something wrong? Why did this happen? Why?' "

These are the questions Kim and I have asked again and again since losing one daughter and watching the other spend four months in the hospital before coming home. Encouraged by Kim's doctors and others not to let the experience scare us from trying again, we are expecting another child in October. Already more than 33 weeks along, Kim has shown no signs of developing pre-eclampsia.

William H. Woodwell Jr. is a freelance writer and editor in Maurertown, Va. He is working on a book about the birth of his daughters.