快猫短视频

BEING born a twin may increase the risk of developing autism. Two separate investigations hint that the competitive environment twins experience in the womb helps to trigger the disorder. However, the scientists are still arguing over the significance of the results, which could turn out to be a mirage.

Many studies appear to show that autism is 70 to 90 per cent attributable to genetic factors. This figure comes from comparing how often autism is shared between genetically identical twins compared with fraternal twins, who are no more alike than any other siblings. But that analysis assumes there is no link between twinning itself and autism.

Now David Greenberg, a geneticist at Columbia University in New York, says this assumption could be false. His doubts surfaced while he was examining a database of families in which at least two siblings had autism. 鈥淚 kept noticing twins in these families,鈥� he says.

When he and his colleagues examined 166 sibling pairs from the database, they found that it contained 17 pairs of identical twins and 12 fraternal ones. That鈥檚 respectively 12 and 4 times the incidence of twins in the general population.

Christopher Gillberg of St George鈥檚 Hospital Medical School in London and his colleagues have found a similar pattern in another database of siblings with autism. Among 79 sibling pairs, they found 9 pairs of identical twins, which is 14 times the level in the general population. They did not see an increased risk for fraternal twins, but that may be because the sample was too small.

There鈥檚 still plenty of evidence that genes play a major role in autism. But if being an identical twin is a stronger risk factor for autism than being a fraternal twin, the twin studies have probably overestimated the influence of genes. It also suggests that an environmental trigger for the disease may occur before birth. Identical twins can be more closely associated in the womb, lacking one or two membranes that separate fraternal twins. This may result in a fiercer competition for resources.

Greenberg speculates that not having enough of these resources, whatever they are, could affect singletons too. 鈥淭he lesson here is that even in the age of the Human Genome Project, you can鈥檛 jump to the conclusion that everything is genetics,鈥� he says.

But Joachim Hallmayer of Stanford University isn鈥檛 convinced by Greenberg鈥檚 and Gillberg鈥檚 studies. When his team examined data from Western Australia on all autism cases, not just siblings, they found only a slight excess of twins in autism families. 鈥淭hat argues against these new results,鈥� he says. He suspects that examining only families with two siblings affected by a disease with a genetic link is likely to recruit more identical twins than normal.

Greenberg counters that if studies of siblings favour the recruitment of identical twins, the same thing should happen with other diseases. But he didn鈥檛 find a striking excess of twins in a database of siblings with insulin-dependent diabetes.