Ask any elementary school teacher about the kid whose foot never stops tapping, whose books and pencils keep falling off his desk, seemingly of their own volition, who shouts and pushes and forgets his lunch, his jacket, his head -- and is impossibly distracted from taking a test by the hum of the overhead lights.

Attention deficit hyperactivity disorder (ADHD) affects three to six percent of school-age children: twice as many boys as girls. And it doesn't go away. Grown-ups learn how to better cope with the disorder, says Jeanette Ramer, associate professor of pediatrics at Penn State's Milton S. Hershey Medical Center, but they remain fidgety and impulsive.

Recent research suggests that ADHD is caused by a chemical imbalance in the brain. Positron emission tomography (PET) scans, able to depict brain metabolism, reveal marked differences in glucose activity in some areas of the ADHD brain. And the majority of ADHD patients are helped substantially by medication: the stimulants Ritalin and Dexedrine, of all things, tend to slow them down. "How," Ramer acknowledges, "is not well known."

One thing that is becoming increasingly clear, however, is that ADHD tends to run in families. About 40 percent of the time, the condition is passed from one generation to the next. Ramer, director of Hershey's ADHD clinic, regularly observes this phenomenon first hand.

"We'll bring a kid in, usually on the recommendation of a teacher," she says. "And we'll ask if anyone else in the family has the same sorts of problems.

"Typically we'll see both parents, and there'll be this sheepish look from one to the other. Or one parent will volunteer. 'I had the same problems when I was in school.'" Then the researchers will ask for old school records and look for comments like "inattentive," "underachiever," or "poor organizational skills." Tell-tale signs.

Ramer and Ellen Hess, assistant professor of neuroscience and anatomy, have constructed pedigrees for six central Pennsylvania families, tracing the occurrence of ADHD over three generations. Establishing the pattern of transmission, Ramer says, is the first step to identifying the genetic link -- the gene or genes responsible for ADHD.

Finding that link will be a complex task. "It's not like a physical disorder, where somebody has it or doesn't," Ramer notes. There is no medical test that can detect ADHD. "There's some question as to whether this is not just one end of a spectrum. All of us have a certain amount of these traits."

On the genetic end, there may well be more than one gene involved in ADHD. Maybe even a combination of genes with specific additive effects, all of which must be present to trigger the condition.

Last year Ramer and Hess, working with statistical geneticist Peter Rogan, thought they had caught an early break in their investigation: a window on ADHD through a strain of mice known as coloboma. "These hyperactive but otherwise functional mice," Ramer explains, "have a genetic deletion in a chromosome that is analogous to chromosome 20 in humans." Curiously, the coloboma mice responded well to Dexedrine, but did not respond to Ritalin, the most commonly used ADHD medication in humans. Ramer thought the particular genetic defect and the response to Dexedrine might be related.

She was wrong. A statistical analysis of five human families with inherited ADHD revealed no chromosome-20 abnormalities that could be linked to the presence of ADHD. "Mice are not men in this situation," Ramer says. The negative result was reported in an article in the American Journal of Medical Genetics.

Undaunted, Ramer and her colleagues have now embarked on a more intensive approach. In collaboration with Maximilian Muenke, a geneticist at Children's Hospital in Philadelphia, they're preparing to scour the entire genome, looking for the genes that link to ADHD in DNA samples provided by Ramer's pedigreed families.

In the lab, Muenke will examine hundreds of markers -- well-identified stretches of human DNA in which there are many normal variations among the general population. He will look for matches among those markers that are common to individuals affected by ADHD. "If you find a region that looks promising," Ramer says, "you can then pinpoint, ultimately getting down to the gene level."

She thinks they'll find more than one ADHD gene, maybe as many as 30. "It could be one gene in some families, another in others.

"Once we know the gene or genes," she adds, "we hope it will help us understand how ADHD works -- what receptors are involved, what's going on with the neurochemistry. This should help us design better medications."

Jeanette C. Ramer, M.D., is associate professor in the division of genetics, department of pediatrics, in the College of Medicine, Milton S. Hershey Medical Center, Pennsylvania State University, 500 University Drive, Hershey PA 17033; 717-531-8006. Ellen J. Hess, Ph.D., is assistant professor of neuroscience and anatomy. Peter K. Rogan, Ph.D., is assistant professor in the division of genetics.