By Rob Stein
Washington Post Staff Writer
Monday, February 9, 2009; A06
In addition to helping thousands of infertile couples have children, "test tube" babies are offering scientists a novel laboratory for resolving one of the most vexing debates in science: nature vs. nurture.
In the first study of its kind, British researchers have studied children conceived through in vitro fertilization (IVF) to examine whether children whose mothers smoked during pregnancy were more likely to develop behavioral problems because of the toxic effects of smoking -- as has been suspected -- or because their mothers passed on a genetic predisposition to antisocial behavior.
The study, which appears to debunk the notion that smoking's effects on the brain of a developing fetus result in antisocial tendencies, could be the first in a series of attempts to use the approach to disentangle whether genes or various prenatal exposures are responsible for later behavioral problems.
"It's incredibly creative," said Avshalom Caspi, a professor of psychiatry and neuroscience at Duke University who was not involved in the research. "It opens up a whole new way of thinking about how to study this problem."
The approach could be applied to a host of questions -- for example, whether women who are stressed while pregnant are more likely to have children who are anxious or depressed, as some studies suggest.
"Many students of human development are interested in trying to figure out just what are the effects of various prenatal conditions, such as exposure to certain toxins," Caspi said. "Is it the genes, or is it the toxin? That is the question."
Researchers have tried to examine these questions using statistical methods, but those approaches are limited. Some have also tried teasing them apart by studying pregnant animals: exposing some to stress, alcohol or cigarette smoke, for example, and examining their offspring. But it is always unclear how well animal studies explain human behavior.
"Ideally, from an experimental point of view, you'd like to expose some infants to smoking and some not, but clearly that's not doable," Caspi said. "It's clearly unethical."
So Anita Thapar, a professor of child and adolescent psychiatry at Cardiff University in Wales, and her colleagues came up with the idea of studying babies born to women undergoing IVF. In some cases, such women are genetically related to their babies because they use their own eggs to create embryos with their husband's sperm. But others have babies who are not genetically related to them because they use eggs or embryos donated by unrelated women. In both cases, the women carry the developing children in their wombs.
Thapar and her colleagues studied 779 children, ages 4 to 10, including about 533 who were genetically related to their mothers and 195 who were not. The researchers examined whether those whose mothers smoked were more likely to have been born underweight and whether they developed behavioral problems, such as temper tantrums or frequent fighting.
"If it's the cigarette smoke having a toxic effect on the brain, it shouldn't matter -- you'd see the relationship with the child's behavior regardless of whether the child is related or unrelated to the mothers who smoke," Thapar said. "But it's not random whether mothers smoke or not. There are a lot of contributing factors, including mothers' personalities. What if it's those factors that are contributing to the behaviors and not the actual smoking?"
As expected, the babies whose mothers smoked were much more likely to be born underweight, regardless of whether they were genetically related to the mothers, the researchers reported last week in the Proceedings of the National Academy of Sciences. That shows a clear biological effect of smoking on birth weight, regardless of genetics. But antisocial behavior was more common only in children who were genetically related to mothers who smoked, indicating that a genetic influence was at work.
"This suggests there may be inherited factors contributing, and it is not the cigarette smoking having the effect," Thapar said. "Other factors, such as a mother's personality traits and other inherited characteristics, are at play during the development of the baby. This says if you find a particular risk factor that looks like prenatal nurture, be careful -- it may really be nature."
Thapar and others stressed, however, that complex behaviors are probably the result of a combination of genes and experience.
"There may be straightforward genetic effects that may be associated with problems like this in children, but it seems more likely that it's probably some kind of combination of genes and environment," said Barbara Maughan, a professor of development epidemiology at King's College London. "It may be what these children are inheriting is an increased tendency to have difficulties if they encounter an adverse environment down the line."
Thapar stressed that the findings in no way suggest it is safe for pregnant women to smoke. Smoking clearly increases the risk of being born underweight, which can increase the risk of a host of medical problems.
"It's important for mothers to quit smoking. Clearly, that will have beneficial impacts on the baby's birth weight," she said. "But in regards to a child's behavior, we need to be careful and think about other types of factors that might be important."
Kenneth Kendler, a professor of psychiatry and human and molecular genetics at Virginia Commonwealth University in Richmond, called the study design "stunning" and "as elegant as they come." But he stressed that the findings should be followed up with additional research, given that in the study the number of women who smoked while carrying genetically unrelated babies was relatively small.
But he agreed that the concept "opens up a new paradigm" for studying similar questions. "Sometimes new technologies in medicine give us new methods to answer questions we could never have answered before," he said. "It's really quite intriguing."