FOR PSYCHOLOGISTS TRYING TO understand how much of someone's appearance, temperament or IQ is the result of the genes he was born with, and how much is the result of parental influence, chance encounters, diet and everything else from the outside world, identical twins separated at birth are more valuable than an unrestricted grant. Identical twins have 100 percent of their genes in common. Separated in infancy, usually through adoption, they are assumed to have 0 percent of their environment in common. But last week researchers led by Bernie Devlin of the University of Pittsburgh reported the results of an experiment that suggests that the ""no shared environment'' assumption is seriously flawed. In fact, twins who have never been in the same place since they popped into the delivery room do, or did, share an environment, argue the scientists. It's called the womb.
Twins studies have come under attack before, especially from scientists who think they artificially inflate the contribution of ""nature'' and lowball that of ""nurture.'' The debate came to a head in 1994 with the publication of ""The Bell Curve'' - in which the authors argue that economic success is explained by IQ and IQ is explained by genes - and has been on a slow boil ever since. Now the Pitt team, prominent critics of that syllogism, has entered the fray. It re- analyzed 212 studies of IQ and genetic inheritance spanning the last 70 years. The mathematical model that best explained the similarities and differences between twins, the team reports in the journal Nature, includes genetic effects and environmental effects, as every other model does - but also what the scientists call the ""maternal effect.'' This effect of the uterine environment accounts for 20 percent of the IQ similarities between twins and 5 percent of the similarities between other siblings, they calculate. ""The effects of genes are correspondingly reduced,'' conclude the authors, to 34 percent rather than the 70 percent calculated from twins studies.
That the uterine environment matters is, at one level, obvious: why else would obstetricians advise pregnant women not to smoke or take illicit drugs, to cut down on alcohol and eat right? And, other things being equal, higher birth weights predict higher IQs. (Put the other way, poor prenatal care among disadvantaged women may partly explain their children's lower IQs and struggles in school.) On the bright side, ""interventions aimed at improving the prenatal environment could lead to a significant increase in the population's IQ,'' says Devlin. For that to happen, of course, researchers would have to identify exactly which aspect of the maternal environment affects a child's intelligence. It could be anything from stress hormones in the mother's blood, to the amount of protein in her diet during the first trimester, to exposure to pollutants during critical periods of fetal brain formation.
Some critics think the Pittsburgh team has not proved the case for a strong effect of the uterine environment on IQ. They agree with the Pitt researchers that fraternal twins raised together are more similar than nontwin siblings raised together, even though nonidentical twins are no more alike genetically than other siblings. ""But it is not at all clear that this is because of the shared intrauterine environment,'' says behavioral geneticist Gregory Carey of the University of Colorado. How, and how strongly, life in the womb affects the child-to-be is only beginning to be understood. But at least it's a step beyond the tired old dyad of genes and environment.