A new study has found that drugs used to treat attention deficit disorders may put kids at risk for of sudden death. But the U.S. Food and Drug Administration is warning the findings should not prompt parents to stop giving the medications to their children.

The study, published Monday in the American Journal of Psychiatry, found there may be an association between stimulant medications used to treat ADHD (attention deficit hyperactivity disorder), such as Ritalin, and sudden death in healthy children.

The drugs already carry warnings about risks of heart attack and stroke in children with underlying heart conditions. But in this new study, researchers found that children who were otherwise healthy and taking the meds were six to seven times more likely to die suddenly for unexplained reasons than those not taking the drug.

The study was partially funded by the FDA, but agency experts said the study methods -- which relied on interviews with parents and physicians years after the children's deaths -- is not the most reliable study method and may have caused errors.

"Since the deaths occurred a long time ago, all of this depended on the memory of people -- relatives and physicians -- involved with the victims," said Dr. Robert Temple, the FDA's director of drug review.

The agency urged parents to discuss safety concerns with their doctor, but to keep children on the treatments.

"FDA believes that this study should not serve as a basis for parents to stop a child's stimulant medication. Parents should discuss concerns about the use of these medicines with the prescribing healthcare professional," the statement said.

Researchers compared 564 children and teens who died of unexplained causes to 564 children and teens who had died in car accidents. The researchers used car accident victims as a comparison group because sudden childhood deaths are rare and difficult to track.

They found that among the unexplained deaths, 10 children were taking an ADHD drug compared with two of the patients killed in car accidents.

The researchers say they're confident the association is real, but the FDA noted the low rate of stimulant use in both study groups may limit the significance of the findings.

The FDA says it is working on a larger, more in-depth study on the potential risks of stimulant medication use in children.

"The data collection for this study will be complete later in 2009," the FDA said.