Still, knowing who may be at risk of autism could improve diagnosis, which might enable earlier intervention.
One study, led by the University of Pennsylvania and Children's Hospital of Philadelphia, followed babies who weighed less than 4.4 pounds at birth through to age 21. Nearly 5 percent of these 623 young adults had an autism-spectrum disorder, five times the rate in the general population.
"Clearly, screening for autism-spectrum disorders in low-birth-weight survivors is critical," said lead author Jennifer Pinto-Martin, a Penn School of Nursing epidemiologist.
Autism includes a host of developmental disorders of varying severity marked by communication problems, compulsive behaviors, and inappropriate social interactions.
Last year, government researchers estimated that in 2006, the disorders affected 1 percent of U.S. children. Though that is in line with rates in other countries, it is 57 percent higher than the U.S. rate in 2002.
Better diagnosis can't fully explain the dramatic increase, experts agree.
In searching for explanations, researchers are looking at lifestyle changes. In recent decades, women have been delaying motherhood, which increases both their chance of needing fertility treatment, and their chance of having a low-birth-weight baby, typically due to prematurity.
These changes have emerged as risk factors for autism:
Two studies presented Thursday linked infertility treatment to the chance of autism. A Harvard University study of 3,985 nurses found that using ovulation-inducing drugs to overcome infertility nearly doubled the odds of having an autistic child. A Tel Aviv University study of 564 autistic children found they were three to four times more likely to have been conceived through in-vitro fertilization and to have been born at very low weights than children in the general population. The mothers of autistic children were also older - 33 on average - compared with 31 for the others.
A Danish study of more than 8,000 children diagnosed with autism found that very low birth weight (less than 3.3 pounds) increased autism risk. However, weight was a stronger risk factor for girls than boys, even though autism is four times more common in boys. Being slightly underweight increased the risk by 50 percent for girls, but not for boys.
A Columbia University study that has followed 108,000 Norwegian infants through age 7 found autism was more common among those whose heads were abnormally small or abnormally large in the first year of life. The researchers said "pediatricians should carefully record" babies' head measurements as a possible indicator.
There were also some reassuring results. A study led by the University of North Carolina concluded that although smoking during pregnancy isn't advisable, it doesn't increase autism risk, as was previously thought.
And the Kennedy Krieger Institute in Baltimore debunked the idea that the stress of having an autistic child drives most parents to divorce. Using data from a 2007 national survey, Krieger researchers found that 64 percent of autistic children live with two married parents - compared with 65 percent for other children.
Contact staff writer Marie McCullough at 215-854-2720 or firstname.lastname@example.org.