African-American children whose mothers drank light to moderate amounts of alcohol during pregnancy had lower IQ scores at age 10 than those born to abstainers, a new study has found.

It also showed that similar drinking habits in Caucasian mothers did not lead to IQ reductions in their children.

The findings suggest a greater vulnerability to prenatal alcohol exposure, perhaps due to genetics, among African Americans, said lead investigator Jennifer Willford, an experimental psychologist at the University of Pittsburgh School of Medicine.

"We know the racial difference was not due to differences in the amount or the pattern of drinking," she said. "And they weren't due to differences in socioeconomic status, either."

In the study, which was published yesterday in the journal Alcoholism: Clinical and Experimental Research, Dr. Willford and her team analyzed information collected from more than 600 pairs of women and children.

The women entered the study as moms-to-be who attended Magee-Womens Hospital's prenatal clinic from 1983 to 1985, and were interviewed several times during pregnancy and after their children were born.

Women who drank an average of less than half a drink per day were categorized as light drinkers, and those who had half a drink to one drink per day were called moderate drinkers. One drink was defined as 12 ounces of beer, 4 ounces of wine, or 1.5 ounces of liquor, Dr. Willford said.

The analysis, which was funded by the National Institute on Alcohol Abuse and Alcoholism and the National Institute on Drug Abuse, focused on expectant mothers with light to moderate drinking patterns. Each child had an IQ test at age 10.

African-American children of moms who drank moderately during the first trimester of pregnancy had scores that were two points lower than those of children of abstainers.

Children of moms who continued to drink in the second trimester had IQ scores nearly five points lower than those of abstainers. Generally, as maternal drinking volume rose, IQ points among their children decreased, Dr. Willford explained.

Prenatal alcohol exposure was not associated with IQ differences in Caucasian children in this analysis, but other studies of the same group showed it led to growth reductions and learning and memory deficits among both the black and white children at age 14.

"There have been a number of other studies that have indicated that there are certain ethnic groups that seem to be more sensitive," said Paul Connor, clinical director of the fetal alcohol and drug unit at the University of Washington's medical school.

But, he cautioned, "no ethnicity is safe from prenatal alcohol exposure."

The Pittsburgh findings are in line with what other researchers have observed, Dr. Connor said.

"Even at low levels of alcohol you can still see cognitive deficits," he said. "They may be very subtle and they may not be picked up clinically, but they can be picked up in large population-based studies."

Dr. Willford is working with neuroimaging techniques to try to figure out what prenatal alcohol exposure does to the brain.

"We've really not identified the neurobiological link for these cognitive deficits," she said.

Dr. Willford advised women who are concerned about their alcohol intake to speak to their physicians, and encouraged pregnant women to abstain from drinking entirely.